/* ============================================================================== US-align: universal structure alignment of monomeric and complex proteins and nucleic acids This program was written by Chengxin Zhang at Yang Zhang lab, Department of Computational Medicine and Bioinformatics, University of Michigan, 100 Washtenaw Ave, Ann Arbor, MI 48109-2218. Please report issues to zhanglab@zhanggroup.org References: * Chengxin Zhang, Morgan Shine, Anna Marie Pyle, Yang Zhang (2022) Nat Methods. 19(9), 1109-1115. * Chengxin Zhang, Anna Marie Pyle (2022) iScience. 25(10), 105218. DISCLAIMER: Permission to use, copy, modify, and distribute this program for any purpose, with or without fee, is hereby granted, provided that the notices on the head, the reference information, and this copyright notice appear in all copies or substantial portions of the Software. It is provided "as is" without express or implied warranty. =============================================================================== ========================= How to install US-align ========================= To compile the program in your Linux computer, simply enter make or g++ -static -O3 -ffast-math -lm -o USalign USalign.cpp The '-static' flag should be removed on Mac OS, which does not support building static executables. USalign compiled on Linux, Mac OS and Linux Subsystem for Windows (WSL2) on Windows 10 onwards can read both uncompressed files and gz compressed files, provided that the "gunzip" command is available. On the other hand, due to the lack of POSIX support on Windows, US-align natively compiled on Windows without WSL2 cannot parse gz compressed files. US-align is known to be compilable by g++ version 4.8.5 or later, clang++ version 12.0.5 or later and mingw-w64 version 9.3 or later. ===================== How to use US-align ===================== You can run the program without arguments to obtain a brief instruction ./USalign structure1.pdb structure2.pdb A full list of available options can be explored by: ./USalign -h */ #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include #include using namespace std; /* pstream.h */ /* the following code block for REDI_PSTREAM_H_SEEN is from Jonathan * Wakely's pstream.h. His original copyright notice is shown below */ // PStreams - POSIX Process I/O for C++ // Copyright (C) 2001 - 2017 Jonathan Wakely // Distributed under the Boost Software License, Version 1.0. // (See accompanying file LICENSE_1_0.txt or copy at // http://www.boost.org/LICENSE_1_0.txt) // /* do not compile on windows, which does not have cygwin */ #if defined(WIN32) || defined(_WIN32) || defined(__WIN32__) && !defined(__CYGWIN__) #define NO_PSTREAM #else #ifndef REDI_PSTREAM_H_SEEN #define REDI_PSTREAM_H_SEEN #include #include #include #include #include #include #include // for min() #include // for errno #include // for size_t, NULL #include // for exit() #include // for pid_t #include // for waitpid() #include // for ioctl() and FIONREAD #if defined(__sun) # include // for FIONREAD on Solaris 2.5 #endif #include // for pipe() fork() exec() and filedes functions #include // for kill() #include // for fcntl() #if REDI_EVISCERATE_PSTREAMS # include // for FILE, fdopen() #endif /// The library version. #define PSTREAMS_VERSION 0x0101 // 1.0.1 /** * @namespace redi * @brief All PStreams classes are declared in namespace redi. * * Like the standard iostreams, PStreams is a set of class templates, * taking a character type and traits type. As with the standard streams * they are most likely to be used with @c char and the default * traits type, so typedefs for this most common case are provided. * * The @c pstream_common class template is not intended to be used directly, * it is used internally to provide the common functionality for the * other stream classes. */ namespace redi { /// Common base class providing constants and typenames. struct pstreams { /// Type used to specify how to connect to the process. typedef std::ios_base::openmode pmode; /// Type used to hold the arguments for a command. typedef std::vector argv_type; /// Type used for file descriptors. typedef int fd_type; static const pmode pstdin = std::ios_base::out; ///< Write to stdin static const pmode pstdout = std::ios_base::in; ///< Read from stdout static const pmode pstderr = std::ios_base::app; ///< Read from stderr /// Create a new process group for the child process. static const pmode newpg = std::ios_base::trunc; protected: enum { bufsz = 32 }; ///< Size of pstreambuf buffers. enum { pbsz = 2 }; ///< Number of putback characters kept. }; /// Class template for stream buffer. template > class basic_pstreambuf : public std::basic_streambuf , public pstreams { public: // Type definitions for dependent types typedef CharT char_type; typedef Traits traits_type; typedef typename traits_type::int_type int_type; typedef typename traits_type::off_type off_type; typedef typename traits_type::pos_type pos_type; /** @deprecated use pstreams::fd_type instead. */ typedef fd_type fd_t; /// Default constructor. basic_pstreambuf(); /// Constructor that initialises the buffer with @a cmd. basic_pstreambuf(const std::string& cmd, pmode mode); /// Constructor that initialises the buffer with @a file and @a argv. basic_pstreambuf( const std::string& file, const argv_type& argv, pmode mode ); /// Destructor. ~basic_pstreambuf(); /// Initialise the stream buffer with @a cmd. basic_pstreambuf* open(const std::string& cmd, pmode mode); /// Initialise the stream buffer with @a file and @a argv. basic_pstreambuf* open(const std::string& file, const argv_type& argv, pmode mode); /// Close the stream buffer and wait for the process to exit. basic_pstreambuf* close(); /// Send a signal to the process. basic_pstreambuf* kill(int signal = SIGTERM); /// Send a signal to the process' process group. basic_pstreambuf* killpg(int signal = SIGTERM); /// Close the pipe connected to the process' stdin. void peof(); /// Change active input source. bool read_err(bool readerr = true); /// Report whether the stream buffer has been initialised. bool is_open() const; /// Report whether the process has exited. bool exited(); #if REDI_EVISCERATE_PSTREAMS /// Obtain FILE pointers for each of the process' standard streams. std::size_t fopen(FILE*& in, FILE*& out, FILE*& err); #endif /// Return the exit status of the process. int status() const; /// Return the error number (errno) for the most recent failed operation. int error() const; protected: /// Transfer characters to the pipe when character buffer overflows. int_type overflow(int_type c); /// Transfer characters from the pipe when the character buffer is empty. int_type underflow(); /// Make a character available to be returned by the next extraction. int_type pbackfail(int_type c = traits_type::eof()); /// Write any buffered characters to the stream. int sync(); /// Insert multiple characters into the pipe. std::streamsize xsputn(const char_type* s, std::streamsize n); /// Insert a sequence of characters into the pipe. std::streamsize write(const char_type* s, std::streamsize n); /// Extract a sequence of characters from the pipe. std::streamsize read(char_type* s, std::streamsize n); /// Report how many characters can be read from active input without blocking. std::streamsize showmanyc(); protected: /// Enumerated type to indicate whether stdout or stderr is to be read. enum buf_read_src { rsrc_out = 0, rsrc_err = 1 }; /// Initialise pipes and fork process. pid_t fork(pmode mode); /// Wait for the child process to exit. int wait(bool nohang = false); /// Return the file descriptor for the output pipe. fd_type& wpipe(); /// Return the file descriptor for the active input pipe. fd_type& rpipe(); /// Return the file descriptor for the specified input pipe. fd_type& rpipe(buf_read_src which); void create_buffers(pmode mode); void destroy_buffers(pmode mode); /// Writes buffered characters to the process' stdin pipe. bool empty_buffer(); bool fill_buffer(bool non_blocking = false); /// Return the active input buffer. char_type* rbuffer(); buf_read_src switch_read_buffer(buf_read_src); private: basic_pstreambuf(const basic_pstreambuf&); basic_pstreambuf& operator=(const basic_pstreambuf&); void init_rbuffers(); pid_t ppid_; // pid of process fd_type wpipe_; // pipe used to write to process' stdin fd_type rpipe_[2]; // two pipes to read from, stdout and stderr char_type* wbuffer_; char_type* rbuffer_[2]; char_type* rbufstate_[3]; /// Index into rpipe_[] to indicate active source for read operations. buf_read_src rsrc_; int status_; // hold exit status of child process int error_; // hold errno if fork() or exec() fails }; /// Class template for common base class. template > class pstream_common : virtual public std::basic_ios , virtual public pstreams { protected: typedef basic_pstreambuf streambuf_type; typedef pstreams::pmode pmode; typedef pstreams::argv_type argv_type; /// Default constructor. pstream_common(); /// Constructor that initialises the stream by starting a process. pstream_common(const std::string& cmd, pmode mode); /// Constructor that initialises the stream by starting a process. pstream_common(const std::string& file, const argv_type& argv, pmode mode); /// Pure virtual destructor. virtual ~pstream_common() = 0; /// Start a process. void do_open(const std::string& cmd, pmode mode); /// Start a process. void do_open(const std::string& file, const argv_type& argv, pmode mode); public: /// Close the pipe. void close(); /// Report whether the stream's buffer has been initialised. bool is_open() const; /// Return the command used to initialise the stream. const std::string& command() const; /// Return a pointer to the stream buffer. streambuf_type* rdbuf() const; #if REDI_EVISCERATE_PSTREAMS /// Obtain FILE pointers for each of the process' standard streams. std::size_t fopen(FILE*& in, FILE*& out, FILE*& err); #endif protected: std::string command_; ///< The command used to start the process. streambuf_type buf_; ///< The stream buffer. }; /** * @class basic_ipstream * @brief Class template for Input PStreams. * * Reading from an ipstream reads the command's standard output and/or * standard error (depending on how the ipstream is opened) * and the command's standard input is the same as that of the process * that created the object, unless altered by the command itself. */ template > class basic_ipstream : public std::basic_istream , public pstream_common , virtual public pstreams { typedef std::basic_istream istream_type; typedef pstream_common pbase_type; using pbase_type::buf_; // declare name in this scope // Ensure a basic_ipstream will read from at least one pipe pmode readable(pmode mode) { if (!(mode & (pstdout|pstderr))) mode |= pstdout; return mode; } public: /// Type used to specify how to connect to the process. typedef typename pbase_type::pmode pmode; /// Type used to hold the arguments for a command. typedef typename pbase_type::argv_type argv_type; /// Default constructor, creates an uninitialised stream. basic_ipstream() : istream_type(NULL), pbase_type() { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ explicit basic_ipstream(const std::string& cmd, pmode mode = pstdout) : istream_type(NULL), pbase_type(cmd, readable(mode)) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ basic_ipstream( const std::string& file, const argv_type& argv, pmode mode = pstdout ) : istream_type(NULL), pbase_type(file, argv, readable(mode)) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling * @c do_open(argv[0],argv,mode|pstdout) * * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ explicit basic_ipstream(const argv_type& argv, pmode mode = pstdout) : istream_type(NULL), pbase_type(argv.at(0), argv, readable(mode)) { } #if __cplusplus >= 201103L template explicit basic_ipstream(std::initializer_list args, pmode mode = pstdout) : basic_ipstream(argv_type(args.begin(), args.end()), mode) { } #endif /** * @brief Destructor. * * Closes the stream and waits for the child to exit. */ ~basic_ipstream() { } /** * @brief Start a process. * * Calls do_open( @a cmd , @a mode|pstdout ). * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ void open(const std::string& cmd, pmode mode = pstdout) { this->do_open(cmd, readable(mode)); } /** * @brief Start a process. * * Calls do_open( @a file , @a argv , @a mode|pstdout ). * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ void open( const std::string& file, const argv_type& argv, pmode mode = pstdout ) { this->do_open(file, argv, readable(mode)); } /** * @brief Set streambuf to read from process' @c stdout. * @return @c *this */ basic_ipstream& out() { this->buf_.read_err(false); return *this; } /** * @brief Set streambuf to read from process' @c stderr. * @return @c *this */ basic_ipstream& err() { this->buf_.read_err(true); return *this; } }; /** * @class basic_opstream * @brief Class template for Output PStreams. * * Writing to an open opstream writes to the standard input of the command; * the command's standard output is the same as that of the process that * created the pstream object, unless altered by the command itself. */ template > class basic_opstream : public std::basic_ostream , public pstream_common , virtual public pstreams { typedef std::basic_ostream ostream_type; typedef pstream_common pbase_type; using pbase_type::buf_; // declare name in this scope public: /// Type used to specify how to connect to the process. typedef typename pbase_type::pmode pmode; /// Type used to hold the arguments for a command. typedef typename pbase_type::argv_type argv_type; /// Default constructor, creates an uninitialised stream. basic_opstream() : ostream_type(NULL), pbase_type() { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ explicit basic_opstream(const std::string& cmd, pmode mode = pstdin) : ostream_type(NULL), pbase_type(cmd, mode|pstdin) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ basic_opstream( const std::string& file, const argv_type& argv, pmode mode = pstdin ) : ostream_type(NULL), pbase_type(file, argv, mode|pstdin) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling * @c do_open(argv[0],argv,mode|pstdin) * * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ explicit basic_opstream(const argv_type& argv, pmode mode = pstdin) : ostream_type(NULL), pbase_type(argv.at(0), argv, mode|pstdin) { } #if __cplusplus >= 201103L /** * @brief Constructor that initialises the stream by starting a process. * * @param args a list of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ template explicit basic_opstream(std::initializer_list args, pmode mode = pstdin) : basic_opstream(argv_type(args.begin(), args.end()), mode) { } #endif /** * @brief Destructor * * Closes the stream and waits for the child to exit. */ ~basic_opstream() { } /** * @brief Start a process. * * Calls do_open( @a cmd , @a mode|pstdin ). * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ void open(const std::string& cmd, pmode mode = pstdin) { this->do_open(cmd, mode|pstdin); } /** * @brief Start a process. * * Calls do_open( @a file , @a argv , @a mode|pstdin ). * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ void open( const std::string& file, const argv_type& argv, pmode mode = pstdin) { this->do_open(file, argv, mode|pstdin); } }; /** * @class basic_pstream * @brief Class template for Bidirectional PStreams. * * Writing to a pstream opened with @c pmode @c pstdin writes to the * standard input of the command. * Reading from a pstream opened with @c pmode @c pstdout and/or @c pstderr * reads the command's standard output and/or standard error. * Any of the process' @c stdin, @c stdout or @c stderr that is not * connected to the pstream (as specified by the @c pmode) * will be the same as the process that created the pstream object, * unless altered by the command itself. */ template > class basic_pstream : public std::basic_iostream , public pstream_common , virtual public pstreams { typedef std::basic_iostream iostream_type; typedef pstream_common pbase_type; using pbase_type::buf_; // declare name in this scope public: /// Type used to specify how to connect to the process. typedef typename pbase_type::pmode pmode; /// Type used to hold the arguments for a command. typedef typename pbase_type::argv_type argv_type; /// Default constructor, creates an uninitialised stream. basic_pstream() : iostream_type(NULL), pbase_type() { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ explicit basic_pstream(const std::string& cmd, pmode mode = pstdout|pstdin) : iostream_type(NULL), pbase_type(cmd, mode) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ basic_pstream( const std::string& file, const argv_type& argv, pmode mode = pstdout|pstdin ) : iostream_type(NULL), pbase_type(file, argv, mode) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling * @c do_open(argv[0],argv,mode) * * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ explicit basic_pstream(const argv_type& argv, pmode mode = pstdout|pstdin) : iostream_type(NULL), pbase_type(argv.at(0), argv, mode) { } #if __cplusplus >= 201103L /** * @brief Constructor that initialises the stream by starting a process. * * @param l a list of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ template explicit basic_pstream(std::initializer_list l, pmode mode = pstdout|pstdin) : basic_pstream(argv_type(l.begin(), l.end()), mode) { } #endif /** * @brief Destructor * * Closes the stream and waits for the child to exit. */ ~basic_pstream() { } /** * @brief Start a process. * * Calls do_open( @a cnd , @a mode ). * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ void open(const std::string& cmd, pmode mode = pstdout|pstdin) { this->do_open(cmd, mode); } /** * @brief Start a process. * * Calls do_open( @a file , @a argv , @a mode ). * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ void open( const std::string& file, const argv_type& argv, pmode mode = pstdout|pstdin ) { this->do_open(file, argv, mode); } /** * @brief Set streambuf to read from process' @c stdout. * @return @c *this */ basic_pstream& out() { this->buf_.read_err(false); return *this; } /** * @brief Set streambuf to read from process' @c stderr. * @return @c *this */ basic_pstream& err() { this->buf_.read_err(true); return *this; } }; /** * @class basic_rpstream * @brief Class template for Restricted PStreams. * * Writing to an rpstream opened with @c pmode @c pstdin writes to the * standard input of the command. * It is not possible to read directly from an rpstream object, to use * an rpstream as in istream you must call either basic_rpstream::out() * or basic_rpstream::err(). This is to prevent accidental reads from * the wrong input source. If the rpstream was not opened with @c pmode * @c pstderr then the class cannot read the process' @c stderr, and * basic_rpstream::err() will return an istream that reads from the * process' @c stdout, and vice versa. * Reading from an rpstream opened with @c pmode @c pstdout and/or * @c pstderr reads the command's standard output and/or standard error. * Any of the process' @c stdin, @c stdout or @c stderr that is not * connected to the pstream (as specified by the @c pmode) * will be the same as the process that created the pstream object, * unless altered by the command itself. */ template > class basic_rpstream : public std::basic_ostream , private std::basic_istream , private pstream_common , virtual public pstreams { typedef std::basic_ostream ostream_type; typedef std::basic_istream istream_type; typedef pstream_common pbase_type; using pbase_type::buf_; // declare name in this scope public: /// Type used to specify how to connect to the process. typedef typename pbase_type::pmode pmode; /// Type used to hold the arguments for a command. typedef typename pbase_type::argv_type argv_type; /// Default constructor, creates an uninitialised stream. basic_rpstream() : ostream_type(NULL), istream_type(NULL), pbase_type() { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ explicit basic_rpstream(const std::string& cmd, pmode mode = pstdout|pstdin) : ostream_type(NULL) , istream_type(NULL) , pbase_type(cmd, mode) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling do_open() with the supplied * arguments. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ basic_rpstream( const std::string& file, const argv_type& argv, pmode mode = pstdout|pstdin ) : ostream_type(NULL), istream_type(NULL), pbase_type(file, argv, mode) { } /** * @brief Constructor that initialises the stream by starting a process. * * Initialises the stream buffer by calling * @c do_open(argv[0],argv,mode) * * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ explicit basic_rpstream(const argv_type& argv, pmode mode = pstdout|pstdin) : ostream_type(NULL), istream_type(NULL), pbase_type(argv.at(0), argv, mode) { } #if __cplusplus >= 201103L /** * @brief Constructor that initialises the stream by starting a process. * * @param l a list of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ template explicit basic_rpstream(std::initializer_list l, pmode mode = pstdout|pstdin) : basic_rpstream(argv_type(l.begin(), l.end()), mode) { } #endif /// Destructor ~basic_rpstream() { } /** * @brief Start a process. * * Calls do_open( @a cmd , @a mode ). * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ void open(const std::string& cmd, pmode mode = pstdout|pstdin) { this->do_open(cmd, mode); } /** * @brief Start a process. * * Calls do_open( @a file , @a argv , @a mode ). * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ void open( const std::string& file, const argv_type& argv, pmode mode = pstdout|pstdin ) { this->do_open(file, argv, mode); } /** * @brief Obtain a reference to the istream that reads * the process' @c stdout. * @return @c *this */ istream_type& out() { this->buf_.read_err(false); return *this; } /** * @brief Obtain a reference to the istream that reads * the process' @c stderr. * @return @c *this */ istream_type& err() { this->buf_.read_err(true); return *this; } }; /// Type definition for common template specialisation. typedef basic_pstreambuf pstreambuf; /// Type definition for common template specialisation. typedef basic_ipstream ipstream; /// Type definition for common template specialisation. typedef basic_opstream opstream; /// Type definition for common template specialisation. typedef basic_pstream pstream; /// Type definition for common template specialisation. typedef basic_rpstream rpstream; /** * When inserted into an output pstream the manipulator calls * basic_pstreambuf::peof() to close the output pipe, * causing the child process to receive the end-of-file indicator * on subsequent reads from its @c stdin stream. * * @brief Manipulator to close the pipe connected to the process' stdin. * @param s An output PStream class. * @return The stream object the manipulator was invoked on. * @see basic_pstreambuf::peof() * @relates basic_opstream basic_pstream basic_rpstream */ template inline std::basic_ostream& peof(std::basic_ostream& s) { typedef basic_pstreambuf pstreambuf_type; if (pstreambuf_type* p = dynamic_cast(s.rdbuf())) p->peof(); return s; } /* * member definitions for pstreambuf */ /** * @class basic_pstreambuf * Provides underlying streambuf functionality for the PStreams classes. */ /** Creates an uninitialised stream buffer. */ template inline basic_pstreambuf::basic_pstreambuf() : ppid_(-1) // initialise to -1 to indicate no process run yet. , wpipe_(-1) , wbuffer_(NULL) , rsrc_(rsrc_out) , status_(-1) , error_(0) { init_rbuffers(); } /** * Initialises the stream buffer by calling open() with the supplied * arguments. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see open() */ template inline basic_pstreambuf::basic_pstreambuf(const std::string& cmd, pmode mode) : ppid_(-1) // initialise to -1 to indicate no process run yet. , wpipe_(-1) , wbuffer_(NULL) , rsrc_(rsrc_out) , status_(-1) , error_(0) { init_rbuffers(); open(cmd, mode); } /** * Initialises the stream buffer by calling open() with the supplied * arguments. * * @param file a string containing the name of a program to execute. * @param argv a vector of argument strings passsed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see open() */ template inline basic_pstreambuf::basic_pstreambuf( const std::string& file, const argv_type& argv, pmode mode ) : ppid_(-1) // initialise to -1 to indicate no process run yet. , wpipe_(-1) , wbuffer_(NULL) , rsrc_(rsrc_out) , status_(-1) , error_(0) { init_rbuffers(); open(file, argv, mode); } /** * Closes the stream by calling close(). * @see close() */ template inline basic_pstreambuf::~basic_pstreambuf() { close(); } /** * Starts a new process by passing @a command to the shell (/bin/sh) * and opens pipes to the process with the specified @a mode. * * If @a mode contains @c pstdout the initial read source will be * the child process' stdout, otherwise if @a mode contains @c pstderr * the initial read source will be the child's stderr. * * Will duplicate the actions of the shell in searching for an * executable file if the specified file name does not contain a slash (/) * character. * * @warning * There is no way to tell whether the shell command succeeded, this * function will always succeed unless resource limits (such as * memory usage, or number of processes or open files) are exceeded. * This means is_open() will return true even if @a command cannot * be executed. * Use pstreambuf::open(const std::string&, const argv_type&, pmode) * if you need to know whether the command failed to execute. * * @param command a string containing a shell command. * @param mode a bitwise OR of one or more of @c out, @c in, @c err. * @return NULL if the shell could not be started or the * pipes could not be opened, @c this otherwise. * @see execl(3) */ template basic_pstreambuf* basic_pstreambuf::open(const std::string& command, pmode mode) { const char * shell_path = "/bin/sh"; #if 0 const std::string argv[] = { "sh", "-c", command }; return this->open(shell_path, argv_type(argv, argv+3), mode); #else basic_pstreambuf* ret = NULL; if (!is_open()) { switch(fork(mode)) { case 0 : // this is the new process, exec command ::execl(shell_path, "sh", "-c", command.c_str(), (char*)NULL); // can only reach this point if exec() failed // parent can get exit code from waitpid() ::_exit(errno); // using std::exit() would make static dtors run twice case -1 : // couldn't fork, error already handled in pstreambuf::fork() break; default : // this is the parent process // activate buffers create_buffers(mode); ret = this; } } return ret; #endif } /** * @brief Helper function to close a file descriptor. * * Inspects @a fd and calls close(3) if it has a non-negative value. * * @param fd a file descriptor. * @relates basic_pstreambuf */ inline void close_fd(pstreams::fd_type& fd) { if (fd >= 0 && ::close(fd) == 0) fd = -1; } /** * @brief Helper function to close an array of file descriptors. * * Calls @c close_fd() on each member of the array. * The length of the array is determined automatically by * template argument deduction to avoid errors. * * @param fds an array of file descriptors. * @relates basic_pstreambuf */ template inline void close_fd_array(pstreams::fd_type (&fds)[N]) { for (std::size_t i = 0; i < N; ++i) close_fd(fds[i]); } /** * Starts a new process by executing @a file with the arguments in * @a argv and opens pipes to the process with the specified @a mode. * * By convention @c argv[0] should be the file name of the file being * executed. * * If @a mode contains @c pstdout the initial read source will be * the child process' stdout, otherwise if @a mode contains @c pstderr * the initial read source will be the child's stderr. * * Will duplicate the actions of the shell in searching for an * executable file if the specified file name does not contain a slash (/) * character. * * Iff @a file is successfully executed then is_open() will return true. * Otherwise, pstreambuf::error() can be used to obtain the value of * @c errno that was set by execvp(3) in the child process. * * The exit status of the new process will be returned by * pstreambuf::status() after pstreambuf::exited() returns true. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode a bitwise OR of one or more of @c out, @c in and @c err. * @return NULL if a pipe could not be opened or if the program could * not be executed, @c this otherwise. * @see execvp(3) */ template basic_pstreambuf* basic_pstreambuf::open( const std::string& file, const argv_type& argv, pmode mode ) { basic_pstreambuf* ret = NULL; if (!is_open()) { // constants for read/write ends of pipe enum { RD, WR }; // open another pipe and set close-on-exec fd_type ck_exec[] = { -1, -1 }; if (-1 == ::pipe(ck_exec) || -1 == ::fcntl(ck_exec[RD], F_SETFD, FD_CLOEXEC) || -1 == ::fcntl(ck_exec[WR], F_SETFD, FD_CLOEXEC)) { error_ = errno; close_fd_array(ck_exec); } else { switch(fork(mode)) { case 0 : // this is the new process, exec command { char** arg_v = new char*[argv.size()+1]; for (std::size_t i = 0; i < argv.size(); ++i) { const std::string& src = argv[i]; char*& dest = arg_v[i]; dest = new char[src.size()+1]; dest[ src.copy(dest, src.size()) ] = '\0'; } arg_v[argv.size()] = NULL; ::execvp(file.c_str(), arg_v); // can only reach this point if exec() failed // parent can get error code from ck_exec pipe error_ = errno; while (::write(ck_exec[WR], &error_, sizeof(error_)) == -1 && errno == EINTR) { } ::close(ck_exec[WR]); ::close(ck_exec[RD]); ::_exit(error_); // using std::exit() would make static dtors run twice } case -1 : // couldn't fork, error already handled in pstreambuf::fork() close_fd_array(ck_exec); break; default : // this is the parent process // check child called exec() successfully ::close(ck_exec[WR]); switch (::read(ck_exec[RD], &error_, sizeof(error_))) { case 0: // activate buffers create_buffers(mode); ret = this; break; case -1: error_ = errno; break; default: // error_ contains error code from child // call wait() to clean up and set ppid_ to 0 this->wait(); break; } ::close(ck_exec[RD]); } } } return ret; } /** * Creates pipes as specified by @a mode and calls @c fork() to create * a new process. If the fork is successful the parent process stores * the child's PID and the opened pipes and the child process replaces * its standard streams with the opened pipes. * * If an error occurs the error code will be set to one of the possible * errors for @c pipe() or @c fork(). * See your system's documentation for these error codes. * * @param mode an OR of pmodes specifying which of the child's * standard streams to connect to. * @return On success the PID of the child is returned in the parent's * context and zero is returned in the child's context. * On error -1 is returned and the error code is set appropriately. */ template pid_t basic_pstreambuf::fork(pmode mode) { pid_t pid = -1; // Three pairs of file descriptors, for pipes connected to the // process' stdin, stdout and stderr // (stored in a single array so close_fd_array() can close all at once) fd_type fd[] = { -1, -1, -1, -1, -1, -1 }; fd_type* const pin = fd; fd_type* const pout = fd+2; fd_type* const perr = fd+4; // constants for read/write ends of pipe enum { RD, WR }; // N.B. // For the pstreambuf pin is an output stream and // pout and perr are input streams. if (!error_ && mode&pstdin && ::pipe(pin)) error_ = errno; if (!error_ && mode&pstdout && ::pipe(pout)) error_ = errno; if (!error_ && mode&pstderr && ::pipe(perr)) error_ = errno; if (!error_) { pid = ::fork(); switch (pid) { case 0 : { // this is the new process // for each open pipe close one end and redirect the // respective standard stream to the other end if (*pin >= 0) { ::close(pin[WR]); ::dup2(pin[RD], STDIN_FILENO); ::close(pin[RD]); } if (*pout >= 0) { ::close(pout[RD]); ::dup2(pout[WR], STDOUT_FILENO); ::close(pout[WR]); } if (*perr >= 0) { ::close(perr[RD]); ::dup2(perr[WR], STDERR_FILENO); ::close(perr[WR]); } #ifdef _POSIX_JOB_CONTROL if (mode&newpg) ::setpgid(0, 0); // Change to a new process group #endif break; } case -1 : { // couldn't fork for some reason error_ = errno; // close any open pipes close_fd_array(fd); break; } default : { // this is the parent process, store process' pid ppid_ = pid; // store one end of open pipes and close other end if (*pin >= 0) { wpipe_ = pin[WR]; ::close(pin[RD]); } if (*pout >= 0) { rpipe_[rsrc_out] = pout[RD]; ::close(pout[WR]); } if (*perr >= 0) { rpipe_[rsrc_err] = perr[RD]; ::close(perr[WR]); } } } } else { // close any pipes we opened before failure close_fd_array(fd); } return pid; } /** * Closes all pipes and calls wait() to wait for the process to finish. * If an error occurs the error code will be set to one of the possible * errors for @c waitpid(). * See your system's documentation for these errors. * * @return @c this on successful close or @c NULL if there is no * process to close or if an error occurs. */ template basic_pstreambuf* basic_pstreambuf::close() { const bool running = is_open(); sync(); // this might call wait() and reap the child process // rather than trying to work out whether or not we need to clean up // just do it anyway, all cleanup functions are safe to call twice. destroy_buffers(pstdin|pstdout|pstderr); // close pipes before wait() so child gets EOF/SIGPIPE close_fd(wpipe_); close_fd_array(rpipe_); do { error_ = 0; } while (wait() == -1 && error() == EINTR); return running ? this : NULL; } /** * Called on construction to initialise the arrays used for reading. */ template inline void basic_pstreambuf::init_rbuffers() { rpipe_[rsrc_out] = rpipe_[rsrc_err] = -1; rbuffer_[rsrc_out] = rbuffer_[rsrc_err] = NULL; rbufstate_[0] = rbufstate_[1] = rbufstate_[2] = NULL; } template void basic_pstreambuf::create_buffers(pmode mode) { if (mode & pstdin) { delete[] wbuffer_; wbuffer_ = new char_type[bufsz]; this->setp(wbuffer_, wbuffer_ + bufsz); } if (mode & pstdout) { delete[] rbuffer_[rsrc_out]; rbuffer_[rsrc_out] = new char_type[bufsz]; rsrc_ = rsrc_out; this->setg(rbuffer_[rsrc_out] + pbsz, rbuffer_[rsrc_out] + pbsz, rbuffer_[rsrc_out] + pbsz); } if (mode & pstderr) { delete[] rbuffer_[rsrc_err]; rbuffer_[rsrc_err] = new char_type[bufsz]; if (!(mode & pstdout)) { rsrc_ = rsrc_err; this->setg(rbuffer_[rsrc_err] + pbsz, rbuffer_[rsrc_err] + pbsz, rbuffer_[rsrc_err] + pbsz); } } } template void basic_pstreambuf::destroy_buffers(pmode mode) { if (mode & pstdin) { this->setp(NULL, NULL); delete[] wbuffer_; wbuffer_ = NULL; } if (mode & pstdout) { if (rsrc_ == rsrc_out) this->setg(NULL, NULL, NULL); delete[] rbuffer_[rsrc_out]; rbuffer_[rsrc_out] = NULL; } if (mode & pstderr) { if (rsrc_ == rsrc_err) this->setg(NULL, NULL, NULL); delete[] rbuffer_[rsrc_err]; rbuffer_[rsrc_err] = NULL; } } template typename basic_pstreambuf::buf_read_src basic_pstreambuf::switch_read_buffer(buf_read_src src) { if (rsrc_ != src) { char_type* tmpbufstate[] = {this->eback(), this->gptr(), this->egptr()}; this->setg(rbufstate_[0], rbufstate_[1], rbufstate_[2]); for (std::size_t i = 0; i < 3; ++i) rbufstate_[i] = tmpbufstate[i]; rsrc_ = src; } return rsrc_; } /** * Suspends execution and waits for the associated process to exit, or * until a signal is delivered whose action is to terminate the current * process or to call a signal handling function. If the process has * already exited (i.e. it is a "zombie" process) then wait() returns * immediately. Waiting for the child process causes all its system * resources to be freed. * * error() will return EINTR if wait() is interrupted by a signal. * * @param nohang true to return immediately if the process has not exited. * @return 1 if the process has exited and wait() has not yet been called. * 0 if @a nohang is true and the process has not exited yet. * -1 if no process has been started or if an error occurs, * in which case the error can be found using error(). */ template int basic_pstreambuf::wait(bool nohang) { int child_exited = -1; if (is_open()) { int exit_status; switch(::waitpid(ppid_, &exit_status, nohang ? WNOHANG : 0)) { case 0 : // nohang was true and process has not exited child_exited = 0; break; case -1 : error_ = errno; break; default : // process has exited ppid_ = 0; status_ = exit_status; child_exited = 1; // Close wpipe, would get SIGPIPE if we used it. destroy_buffers(pstdin); close_fd(wpipe_); // Must free read buffers and pipes on destruction // or next call to open()/close() break; } } return child_exited; } /** * Sends the specified signal to the process. A signal can be used to * terminate a child process that would not exit otherwise. * * If an error occurs the error code will be set to one of the possible * errors for @c kill(). See your system's documentation for these errors. * * @param signal A signal to send to the child process. * @return @c this or @c NULL if @c kill() fails. */ template inline basic_pstreambuf* basic_pstreambuf::kill(int signal) { basic_pstreambuf* ret = NULL; if (is_open()) { if (::kill(ppid_, signal)) error_ = errno; else { #if 0 // TODO call exited() to check for exit and clean up? leave to user? if (signal==SIGTERM || signal==SIGKILL) this->exited(); #endif ret = this; } } return ret; } /** * Sends the specified signal to the process group of the child process. * A signal can be used to terminate a child process that would not exit * otherwise, or to kill the process and its own children. * * If an error occurs the error code will be set to one of the possible * errors for @c getpgid() or @c kill(). See your system's documentation * for these errors. If the child is in the current process group then * NULL will be returned and the error code set to EPERM. * * @param signal A signal to send to the child process. * @return @c this on success or @c NULL on failure. */ template inline basic_pstreambuf* basic_pstreambuf::killpg(int signal) { basic_pstreambuf* ret = NULL; #ifdef _POSIX_JOB_CONTROL if (is_open()) { pid_t pgid = ::getpgid(ppid_); if (pgid == -1) error_ = errno; else if (pgid == ::getpgrp()) error_ = EPERM; // Don't commit suicide else if (::killpg(pgid, signal)) error_ = errno; else ret = this; } #else error_ = ENOTSUP; #endif return ret; } /** * This function can call pstreambuf::wait() and so may change the * object's state if the child process has already exited. * * @return True if the associated process has exited, false otherwise. * @see basic_pstreambuf::wait() */ template inline bool basic_pstreambuf::exited() { return ppid_ == 0 || wait(true)==1; } /** * @return The exit status of the child process, or -1 if wait() * has not yet been called to wait for the child to exit. * @see basic_pstreambuf::wait() */ template inline int basic_pstreambuf::status() const { return status_; } /** * @return The error code of the most recently failed operation, or zero. */ template inline int basic_pstreambuf::error() const { return error_; } /** * Closes the output pipe, causing the child process to receive the * end-of-file indicator on subsequent reads from its @c stdin stream. */ template inline void basic_pstreambuf::peof() { sync(); destroy_buffers(pstdin); close_fd(wpipe_); } /** * Unlike pstreambuf::exited(), this function will not call wait() and * so will not change the object's state. This means that once a child * process is executed successfully this function will continue to * return true even after the process exits (until wait() is called.) * * @return true if a previous call to open() succeeded and wait() has * not been called and determined that the process has exited, * false otherwise. */ template inline bool basic_pstreambuf::is_open() const { return ppid_ > 0; } /** * Toggle the stream used for reading. If @a readerr is @c true then the * process' @c stderr output will be used for subsequent extractions, if * @a readerr is false the the process' stdout will be used. * @param readerr @c true to read @c stderr, @c false to read @c stdout. * @return @c true if the requested stream is open and will be used for * subsequent extractions, @c false otherwise. */ template inline bool basic_pstreambuf::read_err(bool readerr) { buf_read_src src = readerr ? rsrc_err : rsrc_out; if (rpipe_[src]>=0) { switch_read_buffer(src); return true; } return false; } /** * Called when the internal character buffer is not present or is full, * to transfer the buffer contents to the pipe. * * @param c a character to be written to the pipe. * @return @c traits_type::eof() if an error occurs, otherwise if @a c * is not equal to @c traits_type::eof() it will be buffered and * a value other than @c traits_type::eof() returned to indicate * success. */ template typename basic_pstreambuf::int_type basic_pstreambuf::overflow(int_type c) { if (!empty_buffer()) return traits_type::eof(); else if (!traits_type::eq_int_type(c, traits_type::eof())) return this->sputc(c); else return traits_type::not_eof(c); } template int basic_pstreambuf::sync() { return !exited() && empty_buffer() ? 0 : -1; } /** * @param s character buffer. * @param n buffer length. * @return the number of characters written. */ template std::streamsize basic_pstreambuf::xsputn(const char_type* s, std::streamsize n) { std::streamsize done = 0; while (done < n) { if (std::streamsize nbuf = this->epptr() - this->pptr()) { nbuf = std::min(nbuf, n - done); traits_type::copy(this->pptr(), s + done, nbuf); this->pbump(nbuf); done += nbuf; } else if (!empty_buffer()) break; } return done; } /** * @return true if the buffer was emptied, false otherwise. */ template bool basic_pstreambuf::empty_buffer() { const std::streamsize count = this->pptr() - this->pbase(); if (count > 0) { const std::streamsize written = this->write(this->wbuffer_, count); if (written > 0) { if (const std::streamsize unwritten = count - written) traits_type::move(this->pbase(), this->pbase()+written, unwritten); this->pbump(-written); return true; } } return false; } /** * Called when the internal character buffer is is empty, to re-fill it * from the pipe. * * @return The first available character in the buffer, * or @c traits_type::eof() in case of failure. */ template typename basic_pstreambuf::int_type basic_pstreambuf::underflow() { if (this->gptr() < this->egptr() || fill_buffer()) return traits_type::to_int_type(*this->gptr()); else return traits_type::eof(); } /** * Attempts to make @a c available as the next character to be read by * @c sgetc(). * * @param c a character to make available for extraction. * @return @a c if the character can be made available, * @c traits_type::eof() otherwise. */ template typename basic_pstreambuf::int_type basic_pstreambuf::pbackfail(int_type c) { if (this->gptr() != this->eback()) { this->gbump(-1); if (!traits_type::eq_int_type(c, traits_type::eof())) *this->gptr() = traits_type::to_char_type(c); return traits_type::not_eof(c); } else return traits_type::eof(); } template std::streamsize basic_pstreambuf::showmanyc() { int avail = 0; if (sizeof(char_type) == 1) avail = fill_buffer(true) ? this->egptr() - this->gptr() : -1; #ifdef FIONREAD else { if (::ioctl(rpipe(), FIONREAD, &avail) == -1) avail = -1; else if (avail) avail /= sizeof(char_type); } #endif return std::streamsize(avail); } /** * @return true if the buffer was filled, false otherwise. */ template bool basic_pstreambuf::fill_buffer(bool non_blocking) { const std::streamsize pb1 = this->gptr() - this->eback(); const std::streamsize pb2 = pbsz; const std::streamsize npb = std::min(pb1, pb2); char_type* const rbuf = rbuffer(); if (npb) traits_type::move(rbuf + pbsz - npb, this->gptr() - npb, npb); std::streamsize rc = -1; if (non_blocking) { const int flags = ::fcntl(rpipe(), F_GETFL); if (flags != -1) { const bool blocking = !(flags & O_NONBLOCK); if (blocking) ::fcntl(rpipe(), F_SETFL, flags | O_NONBLOCK); // set non-blocking error_ = 0; rc = read(rbuf + pbsz, bufsz - pbsz); if (rc == -1 && error_ == EAGAIN) // nothing available rc = 0; else if (rc == 0) // EOF rc = -1; if (blocking) ::fcntl(rpipe(), F_SETFL, flags); // restore } } else rc = read(rbuf + pbsz, bufsz - pbsz); if (rc > 0 || (rc == 0 && non_blocking)) { this->setg( rbuf + pbsz - npb, rbuf + pbsz, rbuf + pbsz + rc ); return true; } else { this->setg(NULL, NULL, NULL); return false; } } /** * Writes up to @a n characters to the pipe from the buffer @a s. * * @param s character buffer. * @param n buffer length. * @return the number of characters written. */ template inline std::streamsize basic_pstreambuf::write(const char_type* s, std::streamsize n) { std::streamsize nwritten = 0; if (wpipe() >= 0) { nwritten = ::write(wpipe(), s, n * sizeof(char_type)); if (nwritten == -1) error_ = errno; else nwritten /= sizeof(char_type); } return nwritten; } /** * Reads up to @a n characters from the pipe to the buffer @a s. * * @param s character buffer. * @param n buffer length. * @return the number of characters read. */ template inline std::streamsize basic_pstreambuf::read(char_type* s, std::streamsize n) { std::streamsize nread = 0; if (rpipe() >= 0) { nread = ::read(rpipe(), s, n * sizeof(char_type)); if (nread == -1) error_ = errno; else nread /= sizeof(char_type); } return nread; } /** @return a reference to the output file descriptor */ template inline pstreams::fd_type& basic_pstreambuf::wpipe() { return wpipe_; } /** @return a reference to the active input file descriptor */ template inline pstreams::fd_type& basic_pstreambuf::rpipe() { return rpipe_[rsrc_]; } /** @return a reference to the specified input file descriptor */ template inline pstreams::fd_type& basic_pstreambuf::rpipe(buf_read_src which) { return rpipe_[which]; } /** @return a pointer to the start of the active input buffer area. */ template inline typename basic_pstreambuf::char_type* basic_pstreambuf::rbuffer() { return rbuffer_[rsrc_]; } /* * member definitions for pstream_common */ /** * @class pstream_common * Abstract Base Class providing common functionality for basic_ipstream, * basic_opstream and basic_pstream. * pstream_common manages the basic_pstreambuf stream buffer that is used * by the derived classes to initialise an iostream class. */ /** Creates an uninitialised stream. */ template inline pstream_common::pstream_common() : std::basic_ios(NULL) , command_() , buf_() { this->std::basic_ios::rdbuf(&buf_); } /** * Initialises the stream buffer by calling * do_open( @a command , @a mode ) * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, pmode) */ template inline pstream_common::pstream_common(const std::string& cmd, pmode mode) : std::basic_ios(NULL) , command_(cmd) , buf_() { this->std::basic_ios::rdbuf(&buf_); do_open(cmd, mode); } /** * Initialises the stream buffer by calling * do_open( @a file , @a argv , @a mode ) * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see do_open(const std::string&, const argv_type&, pmode) */ template inline pstream_common::pstream_common( const std::string& file, const argv_type& argv, pmode mode ) : std::basic_ios(NULL) , command_(file) , buf_() { this->std::basic_ios::rdbuf(&buf_); do_open(file, argv, mode); } /** * This is a pure virtual function to make @c pstream_common abstract. * Because it is the destructor it will be called by derived classes * and so must be defined. It is also protected, to discourage use of * the PStreams classes through pointers or references to the base class. * * @sa If defining a pure virtual seems odd you should read * http://www.gotw.ca/gotw/031.htm (and the rest of the site as well!) */ template inline pstream_common::~pstream_common() { } /** * Calls rdbuf()->open( @a command , @a mode ) * and sets @c failbit on error. * * @param cmd a string containing a shell command. * @param mode the I/O mode to use when opening the pipe. * @see basic_pstreambuf::open(const std::string&, pmode) */ template inline void pstream_common::do_open(const std::string& cmd, pmode mode) { if (!buf_.open((command_=cmd), mode)) this->setstate(std::ios_base::failbit); } /** * Calls rdbuf()->open( @a file, @a argv, @a mode ) * and sets @c failbit on error. * * @param file a string containing the pathname of a program to execute. * @param argv a vector of argument strings passed to the new program. * @param mode the I/O mode to use when opening the pipe. * @see basic_pstreambuf::open(const std::string&, const argv_type&, pmode) */ template inline void pstream_common::do_open( const std::string& file, const argv_type& argv, pmode mode ) { if (!buf_.open((command_=file), argv, mode)) this->setstate(std::ios_base::failbit); } /** Calls rdbuf->close() and sets @c failbit on error. */ template inline void pstream_common::close() { if (!buf_.close()) this->setstate(std::ios_base::failbit); } /** * @return rdbuf()->is_open(). * @see basic_pstreambuf::is_open() */ template inline bool pstream_common::is_open() const { return buf_.is_open(); } /** @return a string containing the command used to initialise the stream. */ template inline const std::string& pstream_common::command() const { return command_; } /** @return a pointer to the private stream buffer member. */ // TODO document behaviour if buffer replaced. template inline typename pstream_common::streambuf_type* pstream_common::rdbuf() const { return const_cast(&buf_); } #if REDI_EVISCERATE_PSTREAMS /** * @def REDI_EVISCERATE_PSTREAMS * If this macro has a non-zero value then certain internals of the * @c basic_pstreambuf template class are exposed. In general this is * a Bad Thing, as the internal implementation is largely undocumented * and may be subject to change at any time, so this feature is only * provided because it might make PStreams useful in situations where * it is necessary to do Bad Things. */ /** * @warning This function exposes the internals of the stream buffer and * should be used with caution. It is the caller's responsibility * to flush streams etc. in order to clear any buffered data. * The POSIX.1 function fdopen(3) is used to obtain the * @c FILE pointers from the streambuf's private file descriptor * members so consult your system's documentation for * fdopen(3). * * @param in A FILE* that will refer to the process' stdin. * @param out A FILE* that will refer to the process' stdout. * @param err A FILE* that will refer to the process' stderr. * @return An OR of zero or more of @c pstdin, @c pstdout, @c pstderr. * * For each open stream shared with the child process a @c FILE* is * obtained and assigned to the corresponding parameter. For closed * streams @c NULL is assigned to the parameter. * The return value can be tested to see which parameters should be * @c !NULL by masking with the corresponding @c pmode value. * * @see fdopen(3) */ template std::size_t basic_pstreambuf::fopen(FILE*& in, FILE*& out, FILE*& err) { in = out = err = NULL; std::size_t open_files = 0; if (wpipe() > -1) { if ((in = ::fdopen(wpipe(), "w"))) { open_files |= pstdin; } } if (rpipe(rsrc_out) > -1) { if ((out = ::fdopen(rpipe(rsrc_out), "r"))) { open_files |= pstdout; } } if (rpipe(rsrc_err) > -1) { if ((err = ::fdopen(rpipe(rsrc_err), "r"))) { open_files |= pstderr; } } return open_files; } /** * @warning This function exposes the internals of the stream buffer and * should be used with caution. * * @param in A FILE* that will refer to the process' stdin. * @param out A FILE* that will refer to the process' stdout. * @param err A FILE* that will refer to the process' stderr. * @return A bitwise-or of zero or more of @c pstdin, @c pstdout, @c pstderr. * @see basic_pstreambuf::fopen() */ template inline std::size_t pstream_common::fopen(FILE*& fin, FILE*& fout, FILE*& ferr) { return buf_.fopen(fin, fout, ferr); } #endif // REDI_EVISCERATE_PSTREAMS } // namespace redi #endif // REDI_PSTREAM_H_SEEN #endif // WIN32 /* basic_fun.h */ /* File parsing and basic geometry operations */ void PrintErrorAndQuit(const string sErrorString) { cout << sErrorString << endl; exit(1); } template inline T getmin(const T &a, const T &b) { return b void NewArray(A *** array, int Narray1, int Narray2) { *array=new A* [Narray1]; for(int i=0; i void DeleteArray(A *** array, int Narray) { for(int i=0; i &line_vec, const char delimiter=' ') { bool within_word = false; for (size_t pos=0;pos= 0 && idxEnd >= 0) result = inputString.substr(idxBegin, idxEnd + 1 - idxBegin); return result; } size_t get_PDB_lines(const string filename, vector >&PDB_lines, vector &chainID_list, vector &mol_vec, const int ter_opt, const int infmt_opt, const string atom_opt, const bool autojustify, const int split_opt, const int het_opt, const vector&chain2parse, const vector&model2parse) { size_t i=0; // resi i.e. atom index string line; char chainID=0; string resi=""; bool select_atom=false; size_t model_idx=0; vector tmp_str_vec; int compress_type=0; // uncompressed file ifstream fin; #ifndef REDI_PSTREAM_H_SEEN ifstream fin_gz; #else redi::ipstream fin_gz; // if file is compressed if (filename.size()>=3 && filename.substr(filename.size()-3,3)==".gz") { fin_gz.open("gunzip -c '"+filename+"'"); compress_type=1; } else if (filename.size()>=4 && filename.substr(filename.size()-4,4)==".bz2") { fin_gz.open("bzcat '"+filename+"'"); compress_type=2; } else #endif { if (filename=="-") compress_type=-1; else fin.open(filename.c_str()); } if (infmt_opt==0||infmt_opt==-1) // PDB format { map aa3to1; aa3to1[" A"]=aa3to1[" DA"]='a'; aa3to1[" C"]=aa3to1[" DC"]='c'; aa3to1[" G"]=aa3to1[" DG"]='g'; aa3to1[" U"]=aa3to1["PSU"]='u'; aa3to1[" I"]=aa3to1[" DI"]='i'; aa3to1[" T"]='t'; aa3to1["ALA"]='A'; aa3to1["CYS"]='C'; aa3to1["ASP"]='D'; aa3to1["GLU"]='E'; aa3to1["PHE"]='F'; aa3to1["GLY"]='G'; aa3to1["HIS"]='H'; aa3to1["ILE"]='I'; aa3to1["LYS"]='K'; aa3to1["LEU"]='L'; aa3to1["MET"]=aa3to1["MSE"]='M'; aa3to1["ASN"]='N'; aa3to1["PRO"]='P'; aa3to1["GLN"]='Q'; aa3to1["ARG"]='R'; aa3to1["SER"]='S'; aa3to1["THR"]='T'; aa3to1["VAL"]='V'; aa3to1["TRP"]='W'; aa3to1["TYR"]='Y'; aa3to1["ASX"]='B'; aa3to1["GLX"]='Z'; aa3to1["SEC"]='U'; aa3to1["PYL"]='O'; string atom; string resn; string model_index="1"; map alt_id_dict; // resi -> alt_id string resi_chain; while ((compress_type==-1)?cin.good():(compress_type?fin_gz.good():fin.good())) { if (compress_type==-1) getline(cin, line); else if (compress_type) getline(fin_gz, line); else getline(fin, line); if (infmt_opt==-1 && (line.compare(0,5,"loop_")==0 || line.compare(0,1,"#")==0)) // PDBx/mmCIF return get_PDB_lines(filename,PDB_lines,chainID_list, mol_vec, ter_opt, 3, atom_opt, autojustify, split_opt,het_opt, chain2parse, model2parse); if (model2parse.size()) { if (line.size()>=6 && line.compare(0,5,"MODEL")==0) model_index=Trim(line.substr(5,9)); else if (find(model2parse.begin(),model2parse.end(), model_index)==model2parse.end()) continue; } if (i > 0) { if (ter_opt>=1 && line.compare(0,3,"END")==0) break; else if (ter_opt>=3 && line.compare(0,3,"TER")==0) break; } if (line.compare(0,3,"END")==0) { if (split_opt) chainID=0; map ().swap(alt_id_dict); } if (line.size()>=54 && //(line[16]==' ' || line[16]=='A') && ((line.compare(0, 6, "ATOM ")==0) || (line.compare(0, 6, "HETATM")==0 && het_opt==1) || (line.compare(0, 6, "HETATM")==0 && het_opt==2 && line.compare(17,3, "MSE")==0))) { atom=line.substr(12,4); if (autojustify) { resn=line.substr(17,3); if (aa3to1.count(resn)) { atom=Trim(atom); if (atom.size()) { if (atom.size()>=2 && atom[atom.size()-1]=='*') atom=atom.substr(0,atom.size()-1)+"'"; if (atom.size()==1) atom=" "+atom+" "; else if (atom.size()==2) atom=" "+atom+" "; else if (atom.size()==3) atom=" "+atom; } } } if (atom_opt=="auto") { if (line[17]==' ' && (line[18]=='D'||line[18]==' ')) select_atom=(atom==" C3'"); else select_atom=(atom==" CA "); } else if (atom_opt=="PC4'") { if (line[17]==' ' && (line[18]=='D'||line[18]==' ')) select_atom=(atom==" P ")||(atom==" C4'"); else select_atom=(atom==" CA "); } else select_atom=(atom==atom_opt); if (select_atom) { resi_chain=line.substr(21,6); if (alt_id_dict.count(resi_chain)==0) alt_id_dict[resi_chain]=line[16]; else if (alt_id_dict[resi_chain]!=line[16]) continue; if (chain2parse.size() && ( (line[21]==' ' && find(chain2parse.begin(),chain2parse.end(), "_" )==chain2parse.end())|| (line[21]!=' ' && find(chain2parse.begin(), chain2parse.end(), string(1,line[21]))==chain2parse.end()))) continue; if (!chainID) { chainID=line[21]; model_idx++; stringstream i8_stream; i=0; if (split_opt==2) // split by chain { if (chainID==' ') { if (ter_opt>=1) i8_stream << ":_"; else i8_stream<<':'<=1) i8_stream << ':' << chainID; else i8_stream<<':'<=2 && chainID!=line[21]) break; if (split_opt==2 && chainID!=line[21]) { chainID=line[21]; i=0; stringstream i8_stream; if (chainID==' ') { if (ter_opt>=1) i8_stream << ":_"; else i8_stream<<':'<=1) i8_stream << ':' << chainID; else i8_stream<<':'<().swap(aa3to1); map().swap(alt_id_dict); // resi -> alt_id string ().swap(resi_chain); } else if (infmt_opt==1) // SPICKER format { size_t L=0; float x,y,z; stringstream i8_stream; while ((compress_type==-1)?cin.good():(compress_type?fin_gz.good():fin.good())) { if (compress_type==-1) { cin>>L>>x>>y>>z; getline(cin, line); if (!cin.good()) break; } else if (compress_type) { fin_gz>>L>>x>>y>>z; getline(fin_gz, line); if (!fin_gz.good()) break; } else { fin >>L>>x>>y>>z; getline(fin, line); if (!fin.good()) break; } model_idx++; stringstream i8_stream; i8_stream << ':' << model_idx; chainID_list.push_back(i8_stream.str()); PDB_lines.push_back(tmp_str_vec); mol_vec.push_back(0); for (i=0;i>x>>y>>z; else if (compress_type) fin_gz>>x>>y>>z; else fin >>x>>y>>z; i8_stream<<"ATOM "<='a' && line[0]<='z') mol_vec.back()++; // RNA else mol_vec.back()--; } } } else if (infmt_opt==3) // PDBx/mmCIF format { bool loop_ = false; // not reading following content map _atom_site; int atom_site_pos; vector line_vec; string alt_id="."; // alternative location indicator string asym_id="."; // this is similar to chainID, except that // chainID is char while asym_id is a string // with possibly multiple char string prev_asym_id=""; string AA=""; // residue name string atom=""; string prev_resi=""; string model_index=""; // the same as model_idx but type is string stringstream i8_stream; map alt_id_dict; // resi -> alt_id string resi_chain; while ((compress_type==-1)?cin.good():(compress_type?fin_gz.good():fin.good())) { if (compress_type==-1) getline(cin, line); else if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.size()==0) continue; if (loop_) loop_ = (line.size()>=2)?(line.compare(0,2,"# ")):(line.compare(0,1,"#")); if (!loop_) { if (line.compare(0,5,"loop_")) continue; while(1) { if (compress_type==-1) { if (cin.good()) getline(cin, line); else PrintErrorAndQuit("ERROR! Unexpected end of -"); } else if (compress_type) { if (fin_gz.good()) getline(fin_gz, line); else PrintErrorAndQuit("ERROR! Unexpected end of "+filename); } else { if (fin.good()) getline(fin, line); else PrintErrorAndQuit("ERROR! Unexpected end of "+filename); } if (line.size()) break; } if (line.compare(0,11,"_atom_site.")) continue; loop_=true; _atom_site.clear(); atom_site_pos=0; _atom_site[Trim(line.substr(11))]=atom_site_pos; while(1) { if (compress_type==-1) getline(cin, line); else if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.size()==0) continue; if (line.compare(0,11,"_atom_site.")) break; _atom_site[Trim(line.substr(11))]=++atom_site_pos; } if (_atom_site.count("group_PDB")* _atom_site.count("label_atom_id")* _atom_site.count("label_comp_id")* (_atom_site.count("auth_asym_id")+ _atom_site.count("label_asym_id"))* (_atom_site.count("auth_seq_id")+ _atom_site.count("label_seq_id"))* _atom_site.count("Cartn_x")* _atom_site.count("Cartn_y")* _atom_site.count("Cartn_z")==0) { loop_ = false; cerr<<"Warning! Missing one of the following _atom_site data items: group_PDB, label_atom_id, label_comp_id, auth_asym_id/label_asym_id, auth_seq_id/label_seq_id, Cartn_x, Cartn_y, Cartn_z"<=5) continue; AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) continue; if (atom_opt=="auto") { if (AA[0]==' ' && (AA[1]=='D'||AA[1]==' ')) // DNA || RNA select_atom=(atom==" C3'"); else select_atom=(atom==" CA "); } else if (atom_opt=="PC4'") { if (line[17]==' ' && (line[18]=='D'||line[18]==' ')) select_atom=(line.compare(12,4," P ")==0 )||(line.compare(12,4," C4'")==0); else select_atom=(line.compare(12,4," CA ")==0); } else select_atom=(atom==atom_opt); if (!select_atom) continue; if (_atom_site.count("auth_asym_id")) asym_id=line_vec[_atom_site["auth_asym_id"]]; else asym_id=line_vec[_atom_site["label_asym_id"]]; if (asym_id==".") asym_id=" "; if (chain2parse.size() && ( (asym_id==" " && find(chain2parse.begin(),chain2parse.end(), "_" )==chain2parse.end())|| (asym_id!=" " && find(chain2parse.begin(), chain2parse.end(),asym_id )==chain2parse.end()))) continue; if (model2parse.size() && _atom_site.count("pdbx_PDB_model_num") && find(model2parse.begin(), model2parse.end(), line_vec[_atom_site["pdbx_PDB_model_num"]] )==model2parse.end()) continue; if (_atom_site.count("pdbx_PDB_model_num") && model_index!=line_vec[_atom_site["pdbx_PDB_model_num"]]) { model_index=line_vec[_atom_site["pdbx_PDB_model_num"]]; if (PDB_lines.size() && ter_opt>=1) break; if (PDB_lines.size()==0 || split_opt>=1) { PDB_lines.push_back(tmp_str_vec); mol_vec.push_back(0); prev_asym_id=asym_id; if (split_opt==1 && ter_opt==0) chainID_list.push_back( ':'+model_index); else if (split_opt==2 && ter_opt==0) chainID_list.push_back(':'+model_index+','+asym_id); else //if (split_opt==2 && ter_opt==1) chainID_list.push_back(':'+asym_id); //else //chainID_list.push_back(""); } map().swap(alt_id_dict); } if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") resi+=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; else resi+=" "; if (_atom_site.count("label_alt_id")) // in 39.4 % of entries { alt_id=line_vec[_atom_site["label_alt_id"]]; resi_chain=asym_id+resi; if (alt_id_dict.count(resi_chain)==0) alt_id_dict[resi_chain]=alt_id; else if (alt_id_dict.count(resi_chain) && alt_id!=alt_id_dict[resi_chain]) continue; //if (alt_id!="." && alt_id!="A") continue; } if (prev_asym_id!=asym_id) { if (prev_asym_id!="" && ter_opt>=2) break; if (split_opt>=2) { PDB_lines.push_back(tmp_str_vec); mol_vec.push_back(0); if (split_opt==1 && ter_opt==0) chainID_list.push_back( ':'+model_index); else if (split_opt==2 && ter_opt==0) chainID_list.push_back(':'+model_index+','+asym_id); else //if (split_opt==2 && ter_opt==1) chainID_list.push_back(':'+asym_id); //else //chainID_list.push_back(""); } } if (prev_asym_id!=asym_id) prev_asym_id=asym_id; if (AA[0]==' ' && (AA[1]=='D'||AA[1]==' ')) mol_vec.back()++; else mol_vec.back()--; if (prev_resi==resi && atom_opt!="PC4'") cerr<<"Warning! Duplicated residue "<().swap(alt_id_dict); resi_chain.clear(); } if (compress_type>=1) fin_gz.close(); else if (compress_type==0) fin.close(); line.clear(); if (!split_opt) chainID_list.push_back(""); return PDB_lines.size(); } int read_PDB(const vector &PDB_lines, double **a, char *seq, vector &resi_vec, const int read_resi) { size_t i; for (i=0;i=2) resi_vec.push_back(PDB_lines[i].substr(22,5)+ PDB_lines[i][21]); if (read_resi==1) resi_vec.push_back(PDB_lines[i].substr(22,5)); } seq[i]='\0'; return i; } double dist(double x[3], double y[3]) { double d1=x[0]-y[0]; double d2=x[1]-y[1]; double d3=x[2]-y[2]; return (d1*d1 + d2*d2 + d3*d3); } double dot(double *a, double *b) { return (a[0] * b[0] + a[1] * b[1] + a[2] * b[2]); } void transform(double t[3], double u[3][3], double *x, double *x1) { x1[0]=t[0]+dot(&u[0][0], x); x1[1]=t[1]+dot(&u[1][0], x); x1[2]=t[2]+dot(&u[2][0], x); } void do_rotation(double **x, double **x1, int len, double t[3], double u[3][3]) { for(int i=0; i&sequence, const string &fname_lign, const int i_opt) { if (fname_lign == "") PrintErrorAndQuit("Please provide a file name for option -i!"); // open alignment file int n_p = 0;// number of structures in alignment file string line; ifstream fileIn(fname_lign.c_str()); if (fileIn.is_open()) { while (fileIn.good()) { getline(fileIn, line); if (line.compare(0, 1, ">") == 0)// Flag for a new structure { if (n_p >= 2) break; sequence.push_back(""); n_p++; } else if (n_p > 0 && line!="") sequence.back()+=line; } fileIn.close(); } else PrintErrorAndQuit("ERROR! Alignment file does not exist."); if (n_p < 2) PrintErrorAndQuit("ERROR: Fasta format is wrong, two proteins should be included."); if (sequence[0].size() != sequence[1].size()) PrintErrorAndQuit("ERROR! FASTA file is wrong. The length in alignment should be equal for the two aligned proteins."); if (i_opt==3) { int aligned_resNum=0; for (size_t i=0;i&chain_list, const string &name, const string &dir_opt, const string &suffix_opt) { ifstream fp(name.c_str()); if (! fp.is_open()) PrintErrorAndQuit(("Can not open file: "+name+'\n').c_str()); string line; string filename; int a,b; string sep; while (fp.good()) { getline(fp, line); if (! line.size()) continue; line=Trim(line); for (a=0;a<=2;a++) { if (a==0) sep=""; else if (a==1) sep="/"; else if (a==2) sep="\\"; filename=dir_opt+sep+line+suffix_opt; if (isfile(filename)) break; if (suffix_opt.size()) { filename=dir_opt+sep+line; if (isfile(filename)) break; } else { filename=dir_opt+sep+line+".pdb"; if (isfile(filename)) break; filename=dir_opt+sep+line+".cif"; if (isfile(filename)) break; } filename.clear(); } if (filename.size()==0) { filename=dir_opt+line+suffix_opt; cerr<<"WARNING! "<&chain1_list, vector&chain2_list, const string &name, const string &dirpair_opt, const string &suffix_opt) { ifstream fp(name.c_str()); if (! fp.is_open()) PrintErrorAndQuit(("Can not open file: "+name+'\n').c_str()); string line; string filename; int a,b; size_t i; string sep,filename1,filename2; vector line_vec; while (fp.good()) { getline(fp, line); if (! line.size()) continue; line=Trim(line); split(line, line_vec, '\t'); if (line_vec.size()==2) { filename1=line_vec[0]; filename2=line_vec[1]; for (i=0;i<2;i++) line_vec[i].clear(); line_vec.clear(); } else { for (i=0;i ? @ A B C D E F G H I J K L M N O P Q R S T U V W X Y Z [ \ ] ^ _ ` a b c d e f g h i j k l m n o p q r s t u v w x y z { | } ~ DEL { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//0 '\0' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//1 SOH { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//2 STX { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//3 ETX { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//4 EOT { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//5 ENQ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//6 ACK { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//7 '\a' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//8 '\b' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//9 '\t' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//10 '\n' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//11 '\v' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//12 '\f' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//13 '\r' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//14 SO { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//15 SI { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//16 DLE { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//17 DC1 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//18 DC2 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//19 DC3 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//20 DC4 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//21 NAK { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//22 SYN { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//23 ETB { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//24 CAN { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//25 EM { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//26 SUB { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//27 ESC { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//28 FS { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//29 GS { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//30 RS { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//31 US { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//32 ' ' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//33 ! { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//34 " { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//35 # { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//36 $ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//37 % { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//38 & { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//39 ' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//40 ( { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//41 ) { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4,-4,-4,-4,-4,-4,-4,-4,-4, 0,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//42 * { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//43 + { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//44 , { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//45 - { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//46 . { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//47 / { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//48 0 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//49 1 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//50 2 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//51 3 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//52 4 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//53 5 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//54 6 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//55 7 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//56 8 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//57 9 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//58 : { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//59 ; { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//60 < { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//61 = { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//62 > { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//63 ? { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//64 @ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 4,-2, 0,-2,-1,-2, 0,-2,-1, 0,-1,-1,-1,-2,-1,-1,-1,-1, 1, 0, 0, 0,-3, 0,-2,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//65 A { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2, 4,-3, 4, 1,-3,-1, 0,-3, 0, 0,-4,-3, 3, 0,-2, 0,-1, 0,-1,-3,-3,-4,-1,-3, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//66 B { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 9,-3,-4,-2,-3,-3,-1, 0,-3,-1,-1,-3,-3,-3,-3,-3,-1,-1, 9,-1,-2,-2,-2,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//67 C { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2, 4,-3, 6, 2,-3,-1,-1,-3, 0,-1,-4,-3, 1,-1,-1, 0,-2, 0,-1,-3,-3,-4,-1,-3, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//68 D { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1, 1,-4, 2, 5,-3,-2, 0,-3, 0, 1,-3,-2, 0, 1,-1, 2, 0, 0,-1,-4,-2,-3,-1,-2, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//69 E { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2,-3,-2,-3,-3, 6,-3,-1, 0, 0,-3, 0, 0,-3,-3,-4,-3,-3,-2,-2,-2,-1, 1,-1, 3,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//70 F { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-3,-1,-2,-3, 6,-2,-4, 0,-2,-4,-3, 0,-2,-2,-2,-2, 0,-2,-3,-3,-2,-1,-3,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//71 G { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2, 0,-3,-1, 0,-1,-2, 8,-3, 0,-1,-3,-2, 1,-1,-2, 0, 0,-1,-2,-3,-3,-2,-1, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//72 H { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-3,-1,-3,-3, 0,-4,-3, 4, 0,-3, 2, 1,-3,-3,-3,-3,-3,-2,-1,-1, 3,-3,-1,-1,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//73 I { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//74 J { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1, 0,-3,-1, 1,-3,-2,-1,-3, 0, 5,-2,-1, 0, 5,-1, 1, 2, 0,-1,-3,-2,-3,-1,-2, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//75 K { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-4,-1,-4,-3, 0,-4,-3, 2, 0,-2, 4, 2,-3,-2,-3,-2,-2,-2,-1,-1, 1,-2,-1,-1,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//76 L { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-3,-1,-3,-2, 0,-3,-2, 1, 0,-1, 2, 5,-2,-1,-2, 0,-1,-1,-1,-1, 1,-1,-1,-1,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//77 M { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2, 3,-3, 1, 0,-3, 0, 1,-3, 0, 0,-3,-2, 6, 0,-2, 0, 0, 1, 0,-3,-3,-4,-1,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//78 N { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1, 0,-3,-1, 1,-3,-2,-1,-3, 0, 5,-2,-1, 0, 5,-1, 1, 2, 0,-1,-3,-2,-3,-1,-2, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//79 O { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-2,-3,-1,-1,-4,-2,-2,-3, 0,-1,-3,-2,-2,-1, 7,-1,-2,-1,-1,-3,-2,-4,-2,-3,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//80 P { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1, 0,-3, 0, 2,-3,-2, 0,-3, 0, 1,-2, 0, 0, 1,-1, 5, 1, 0,-1,-3,-2,-2,-1,-1, 3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//81 Q { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-1,-3,-2, 0,-3,-2, 0,-3, 0, 2,-2,-1, 0, 2,-2, 1, 5,-1,-1,-3,-3,-3,-1,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//82 R { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,-1, 0, 0,-2, 0,-1,-2, 0, 0,-2,-1, 1, 0,-1, 0,-1, 4, 1,-1,-2,-3, 0,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//83 S { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-1,-1,-1,-2,-2,-2,-1, 0,-1,-1,-1, 0,-1,-1,-1,-1, 1, 5,-1, 0,-2, 0,-2,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//84 T { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 9,-3,-4,-2,-3,-3,-1, 0,-3,-1,-1,-3,-3,-3,-3,-3,-1,-1, 9,-1,-2,-2,-2,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//85 U { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3,-1,-3,-2,-1,-3,-3, 3, 0,-2, 1, 1,-3,-2,-2,-2,-3,-2, 0,-1, 4,-3,-1,-1,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//86 V { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3,-4,-2,-4,-3, 1,-2,-2,-3, 0,-3,-2,-1,-4,-3,-4,-2,-3,-3,-2,-2,-3,11,-2, 2,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//87 W { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1,-2,-1,-1,-1,-1,-1,-1, 0,-1,-1,-1,-1,-1,-2,-1,-1, 0, 0,-2,-1,-2,-1,-1,-1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//88 X { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-2,-3,-2,-3,-2, 3,-3, 2,-1, 0,-2,-1,-1,-2,-2,-3,-1,-2,-2,-2,-2,-1, 2,-1, 7,-2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//89 Y { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-1, 1,-3, 1, 4,-3,-2, 0,-3, 0, 1,-3,-1, 0, 1,-1, 3, 0, 0,-1,-3,-2,-3,-1,-2, 4, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//90 Z { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//91 [ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//92 '\' { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//93 ] { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//94 ^ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//95 _ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//96 ` { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 0,-3, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//97 a { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//98 b { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0, 2, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//99 c { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//100 d { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//101 e { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//102 f { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0,-3, 0, 0, 0, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//103 g { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//104 h { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//105 i { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//106 j { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//107 k { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//108 l { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//109 m { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//110 n { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//111 o { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//112 p { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//113 q { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//114 r { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//115 s { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0,-3, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//116 t { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,-3, 0,-3, 0, 0, 0,-3, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 2, 2, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//117 u { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//118 v { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//119 w { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//120 x { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//121 y { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//122 z { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//123 { { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//124 | { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//125 } { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//126 ~ { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0},//127 DEL }; #define MAX(A,B) ((A)>(B)?(A):(B)) const int gapopen_blosum62=-11; const int gapext_blosum62=-1; const int gapopen_blastn=-15; //-5; const int gapext_blastn =-4; //-2; /* initialize matrix in gotoh algorithm */ void init_gotoh_mat(int **S, int **JumpH, int **JumpV, int **P, int **H, int **V, const int xlen, const int ylen, const int gapopen, const int gapext, const int glocal=0, const int alt_init=1) { // fill first row/colum of JumpH,jumpV and path matrix P int i,j; for (i=0;i=aln_score) { max_aln_i=i; max_aln_j=j; aln_score=S[i][j]; } } } // reset all path after [max_aln_i][max_aln_j] for (i=max_aln_i+1;i0 * 1 : \ match-mismatch * 2 : | vertical gap (insertion) * 4 : - horizontal gap (deletion) * JumpH - horizontal long gap number. * JumpV - vertical long gap number. * all matrices are in the size of [len(seqx)+1]*[len(seqy)+1] * * glocal - global or local alignment * 0 : global alignment (Needleman-Wunsch dynamic programming) * 1 : glocal-query alignment * 2 : glocal-both alignment * 3 : local alignment (Smith-Waterman dynamic programming) * * alt_init - whether to adopt alternative matrix initialization * 1 : use wei zheng's matrix initialization * 0 : use yang zhang's matrix initialization, does NOT work * for glocal alignment */ int calculate_score_gotoh(const int xlen,const int ylen, int **S, int** JumpH, int** JumpV, int **P, const int gapopen,const int gapext, const int glocal=0, const int alt_init=1) { int **H; int **V; NewArray(&H,xlen+1,ylen+1); // penalty score for horizontal long gap NewArray(&V,xlen+1,ylen+1); // penalty score for vertical long gap // fill first row/colum of JumpH,jumpV and path matrix P int i,j; init_gotoh_mat(S, JumpH, JumpV, P, H, V, xlen, ylen, gapopen, gapext, glocal, alt_init); // fill S and P int diag_score,left_score,up_score; for (i=1;i=3) { H[i][j]=MAX(S[i][j-1]+gapopen,H[i][j-1]+gapext); JumpH[i][j]=(H[i][j]==H[i][j-1]+gapext)?(JumpH[i][j-1]+1):1; } else { H[i][j]=MAX(S[i][j-1],H[i][j-1]); JumpH[i][j]=(H[i][j]==H[i][j-1])?(JumpH[i][j-1]+1):1; } // penalty of consective insertion if (glocal<2 || j=3) { V[i][j]=MAX(S[i-1][j]+gapopen,V[i-1][j]+gapext); JumpV[i][j]=(V[i][j]==V[i-1][j]+gapext)?(JumpV[i-1][j]+1):1; } else { V[i][j]=MAX(S[i-1][j],V[i-1][j]); JumpV[i][j]=(V[i][j]==V[i-1][j])?(JumpV[i-1][j]+1):1; } diag_score=S[i-1][j-1]+S[i][j]; // match-mismatch '\' left_score=H[i][j]; // deletion '-' up_score =V[i][j]; // insertion '|' if (diag_score>=left_score && diag_score>=up_score) { S[i][j]=diag_score; P[i][j]+=1; } if (up_score>=diag_score && up_score>=left_score) { S[i][j]=up_score; P[i][j]+=2; } if (left_score>=diag_score && left_score>=up_score) { S[i][j]=left_score; P[i][j]+=4; } if (glocal>=3 && S[i][j]<0) { S[i][j]=0; P[i][j]=0; H[i][j]=0; V[i][j]=0; JumpH[i][j]=0; JumpV[i][j]=0; } } } int aln_score=S[xlen][ylen]; // re-fill first row/column of path matrix P for back-tracing for (i=1;i0) P[i][0]=2; // | for (j=1;j0) P[0][j]=4; // - // calculate alignment score and alignment path for swalign if (glocal>=3) find_highest_align_score(S,P,aln_score,xlen,ylen); // release memory DeleteArray(&H,xlen+1); DeleteArray(&V,xlen+1); return aln_score; // final alignment score } /* trace back dynamic programming path to diciper pairwise alignment */ void trace_back_gotoh(const char *seqx, const char *seqy, int ** JumpH, int ** JumpV, int ** P, string& seqxA, string& seqyA, const int xlen, const int ylen, int *invmap, const int invmap_only=1) { int i,j; int gaplen,p; char *buf=NULL; if (invmap_only) for (j = 0; j < ylen; j++) invmap[j] = -1; if (invmap_only!=1) buf=new char [MAX(xlen,ylen)+1]; i=xlen; j=ylen; while(i+j) { gaplen=0; if (P[i][j]>=4) { gaplen=JumpH[i][j]; j-=gaplen; if (invmap_only==1) continue; strncpy(buf,seqy+j,gaplen); buf[gaplen]=0; seqyA=buf+seqyA; for (p=0;p= 2) { gaplen=JumpV[i][j]; i-=gaplen; if (invmap_only==1) continue; strncpy(buf,seqx+i,gaplen); buf[gaplen]=0; seqxA=buf+seqxA; for (p=0;p=0;i--) { for (j=ylen;j>=0;j--) { if (P[i][j]!=0) { found_start_cell=true; break; } } if (found_start_cell) break; } /* copy C terminal sequence */ if (invmap_only!=1) { for (p=0;p=4) { gaplen=JumpH[i][j]; j-=gaplen; if (invmap_only==1) continue; strncpy(buf,seqy+j,gaplen); buf[gaplen]=0; seqyA=buf+seqyA; for (p=0;p= 2) { gaplen=JumpV[i][j]; i-=gaplen; if (invmap_only==1) continue; strncpy(buf,seqx+i,gaplen); buf[gaplen]=0; seqxA=buf+seqxA; for (p=0;p0) // RNA or DNA { gapopen=gapopen_blastn; gapext =gapext_blastn; if (glocal==3) { gapopen=-5; gapext =-2; } } for (i=0;i &sequence, char *seqx, char *seqy, const vector resi_vec1, const vector resi_vec2, const int byresi_opt) { sequence.clear(); sequence.push_back(""); sequence.push_back(""); int i1=0; // positions in resi_vec1 int i2=0; // positions in resi_vec2 int xlen=resi_vec1.size(); int ylen=resi_vec2.size(); if (byresi_opt==4 || byresi_opt==5 || byresi_opt==7) // global or glocal sequence alignment { int *invmap; int glocal=0; if (byresi_opt==5 || byresi_opt==7) glocal=2; int mol_type=0; for (i1=0;i1 chainID_map1; map chainID_map2; if (byresi_opt==3) { vector chainID_vec; string chainID; stringstream ss; int i; for (i=0;i().swap(chainID_vec); } string chainID1=""; string chainID2=""; string chainID1_prev=""; string chainID2_prev=""; while(i1 atoi(resi_vec2[i2].substr(0,4).c_str())) { sequence[0]+='-'; sequence[1]+=seqy[i2++]; } else { sequence[0]+=seqx[i1++]; sequence[1]+=seqy[i2++]; } chainID1_prev=chainID1; chainID2_prev=chainID2; } else { if (chainID1_prev==chainID1 && chainID2_prev!=chainID2) { sequence[0]+=seqx[i1++]; sequence[1]+='-'; chainID1_prev=chainID1; } else if (chainID1_prev!=chainID1 && chainID2_prev==chainID2) { sequence[0]+='-'; sequence[1]+=seqy[i2++]; chainID2_prev=chainID2; } else { sequence[0]+=seqx[i1++]; sequence[1]+=seqy[i2++]; chainID1_prev=chainID1; chainID2_prev=chainID2; } } } map().swap(chainID_map1); map().swap(chainID_map2); chainID1.clear(); chainID2.clear(); chainID1_prev.clear(); chainID2_prev.clear(); return sequence[0].size(); } /* extract pairwise sequence alignment from residue index vectors, * return length of alignment, including gap. */ int extract_aln_from_resi(vector &sequence, char *seqx, char *seqy, const vector resi_vec1, const vector resi_vec2, const vector xlen_vec, const vector ylen_vec, const int chain_i, const int chain_j, const int byresi_opt) { sequence.clear(); sequence.push_back(""); sequence.push_back(""); int i1=0; // positions in resi_vec1 int i2=0; // positions in resi_vec2 int xlen=xlen_vec[chain_i]; int ylen=ylen_vec[chain_j]; int i,j; for (i=0;i atoi(resi_vec2[j+i2].substr(0,4).c_str())) { sequence[0]+='-'; sequence[1]+=seqy[j++]; } else { sequence[0]+=seqx[i++]; sequence[1]+=seqy[j++]; } } if (i4.25 Lnorm=getmin(xlen, ylen); //normalize TMscore by this in searching if (Lnorm<=19) //update 15-->19 d0=0.168; //update 0.5-->0.168 else d0=(1.24*pow((Lnorm*1.0-15), 1.0/3)-1.8); D0_MIN=d0+0.8; //this should be moved to above d0=D0_MIN; //update: best for search d0_search=d0; if (d0_search>8) d0_search=8; if (d0_search<4.5) d0_search=4.5; score_d8=1.5*pow(Lnorm*1.0, 0.3)+3.5; //remove pairs with dis>d8 during search & final } void parameter_set4final_C3prime(const double len, double &D0_MIN, double &Lnorm, double &d0, double &d0_search) { D0_MIN=0.3; Lnorm=len; //normalize TMscore by this in searching if(Lnorm<=11) d0=0.3; else if(Lnorm>11&&Lnorm<=15) d0=0.4; else if(Lnorm>15&&Lnorm<=19) d0=0.5; else if(Lnorm>19&&Lnorm<=23) d0=0.6; else if(Lnorm>23&&Lnorm<30) d0=0.7; else d0=(0.6*pow((Lnorm*1.0-0.5), 1.0/2)-2.5); d0_search=d0; if (d0_search>8) d0_search=8; if (d0_search<4.5) d0_search=4.5; } void parameter_set4final(const double len, double &D0_MIN, double &Lnorm, double &d0, double &d0_search, const int mol_type) { if (mol_type>0) // RNA { parameter_set4final_C3prime(len, D0_MIN, Lnorm, d0, d0_search); return; } D0_MIN=0.5; Lnorm=len; //normalize TMscore by this in searching if (Lnorm<=21) d0=0.5; else d0=(1.24*pow((Lnorm*1.0-15), 1.0/3)-1.8); if (d08) d0_search=8; if (d0_search<4.5) d0_search=4.5; } void parameter_set4scale(const int len, const double d_s, double &Lnorm, double &d0, double &d0_search) { d0=d_s; Lnorm=len; //normalize TMscore by this in searching d0_search=d0; if (d0_search>8) d0_search=8; if (d0_search<4.5) d0_search=4.5; } /* NW.h */ /* Partial implementation of Needleman-Wunsch (NW) dynamic programming for * global alignment. The three NWDP_TM functions below are not complete * implementation of NW algorithm because gap jumping in the standard Gotoh * algorithm is not considered. Since the gap opening and gap extension is * the same, this is not a problem. This code was exploited in TM-align * because it is about 1.5 times faster than a complete NW implementation. * Nevertheless, if gap opening != gap extension shall be implemented in * the future, the Gotoh algorithm must be implemented. In rare scenarios, * it is also possible to have asymmetric alignment (i.e. * TMalign A.pdb B.pdb and TMalign B.pdb A.pdb have different TM_A and TM_B * values) caused by the NWPD_TM implement. */ /* Input: score[1:len1, 1:len2], and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_TM(double **score, bool **path, double **val, int len1, int len2, double gap_open, int j2i[]) { int i, j; double h, v, d; //initialization for(i=0; i<=len1; i++) { val[i][0]=0; //val[i][0]=i*gap_open; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { val[0][j]=0; //val[0][j]=j*gap_open; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } //decide matrix and path for(i=1; i<=len1; i++) { for(j=1; j<=len2; j++) { d=val[i-1][j-1]+score[i][j]; //diagonal //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } /* Input: vectors x, y, rotation matrix t, u, scale factor d02, and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_TM(bool **path, double **val, double **x, double **y, int len1, int len2, double t[3], double u[3][3], double d02, double gap_open, int j2i[]) { int i, j; double h, v, d; //initialization. use old val[i][0] and val[0][j] initialization //to minimize difference from TMalign fortran version for(i=0; i<=len1; i++) { val[i][0]=0; //val[i][0]=i*gap_open; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { val[0][j]=0; //val[0][j]=j*gap_open; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } double xx[3], dij; //decide matrix and path for(i=1; i<=len1; i++) { transform(t, u, &x[i-1][0], xx); for(j=1; j<=len2; j++) { dij=dist(xx, &y[j-1][0]); d=val[i-1][j-1] + 1.0/(1+dij/d02); //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } /* This is the same as the previous NWDP_TM, except for the lack of rotation * Input: vectors x, y, scale factor d02, and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_SE(bool **path, double **val, double **x, double **y, int len1, int len2, double d02, double gap_open, int j2i[]) { int i, j; double h, v, d; for(i=0; i<=len1; i++) { val[i][0]=0; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { val[0][j]=0; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } double dij; //decide matrix and path for(i=1; i<=len1; i++) { for(j=1; j<=len2; j++) { dij=dist(&x[i-1][0], &y[j-1][0]); d=val[i-1][j-1] + 1.0/(1+dij/d02); //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } void NWDP_SE(bool **path, double **val, double **x, double **y, int len1, int len2, double d02, double gap_open, int j2i[], const int hinge) { if (hinge==0) { NWDP_SE(path, val, x, y, len1, len2, d02, gap_open, j2i); return; } int i, j; double h, v, d; int L=(len2>len1)?len2:len1; int int_min=L*(gap_open-1); for (i=0; i<=len1; i++) { for (j=0; j<=len2; j++) { val[i][j]=0; path[i][j]=false; } } /* fill in old j2i */ int k=0; for (j=0; j=h && d>=v && val[i][j]==0) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment for (j=0;j<=len2;j++) j2i[j]=-1; i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } /* +ss * Input: secondary structure secx, secy, and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_TM(bool **path, double **val, const char *secx, const char *secy, const int len1, const int len2, const double gap_open, int j2i[]) { int i, j; double h, v, d; //initialization for(i=0; i<=len1; i++) { val[i][0]=0; //val[i][0]=i*gap_open; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { val[0][j]=0; //val[0][j]=j*gap_open; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } //decide matrix and path for(i=1; i<=len1; i++) { for(j=1; j<=len2; j++) { d=val[i-1][j-1] + 1.0*(secx[i-1]==secy[j-1]); //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } /* Kabsch.h */ /************************************************************************** Implemetation of Kabsch algoritm for finding the best rotation matrix --------------------------------------------------------------------------- x - x(i,m) are coordinates of atom m in set x (input) y - y(i,m) are coordinates of atom m in set y (input) n - n is number of atom pairs (input) mode - 0:calculate rms only (input) 1:calculate u,t only (takes medium) 2:calculate rms,u,t (takes longer) rms - sum of w*(ux+t-y)**2 over all atom pairs (output) u - u(i,j) is rotation matrix for best superposition (output) t - t(i) is translation vector for best superposition (output) **************************************************************************/ bool Kabsch(double **x, double **y, int n, int mode, double *rms, double t[3], double u[3][3]) { int i, j, m, m1, l, k; double e0, rms1, d, h, g; double cth, sth, sqrth, p, det, sigma; double xc[3], yc[3]; double a[3][3], b[3][3], r[3][3], e[3], rr[6], ss[6]; double sqrt3 = 1.73205080756888, tol = 0.01; int ip[] = { 0, 1, 3, 1, 2, 4, 3, 4, 5 }; int ip2312[] = { 1, 2, 0, 1 }; int a_failed = 0, b_failed = 0; double epsilon = 0.00000001; //initialization *rms = 0; rms1 = 0; e0 = 0; double c1[3], c2[3]; double s1[3], s2[3]; double sx[3], sy[3], sz[3]; for (i = 0; i < 3; i++) { s1[i] = 0.0; s2[i] = 0.0; sx[i] = 0.0; sy[i] = 0.0; sz[i] = 0.0; } for (i = 0; i<3; i++) { xc[i] = 0.0; yc[i] = 0.0; t[i] = 0.0; for (j = 0; j<3; j++) { u[i][j] = 0.0; r[i][j] = 0.0; a[i][j] = 0.0; if (i == j) { u[i][j] = 1.0; a[i][j] = 1.0; } } } if (n<1) return false; //compute centers for vector sets x, y for (i = 0; i0) { d = spur*spur; h = d - cof; g = (spur*cof - det) / 2.0 - spur*h; if (h>0) { sqrth = sqrt(h); d = h*h*h - g*g; if (d<0.0) d = 0.0; d = atan2(sqrt(d), -g) / 3.0; cth = sqrth * cos(d); sth = sqrth*sqrt3*sin(d); e[0] = (spur + cth) + cth; e[1] = (spur - cth) + sth; e[2] = (spur - cth) - sth; if (mode != 0) {//compute a for (l = 0; l<3; l = l + 2) { d = e[l]; ss[0] = (d - rr[2]) * (d - rr[5]) - rr[4] * rr[4]; ss[1] = (d - rr[5]) * rr[1] + rr[3] * rr[4]; ss[2] = (d - rr[0]) * (d - rr[5]) - rr[3] * rr[3]; ss[3] = (d - rr[2]) * rr[3] + rr[1] * rr[4]; ss[4] = (d - rr[0]) * rr[4] + rr[1] * rr[3]; ss[5] = (d - rr[0]) * (d - rr[2]) - rr[1] * rr[1]; if (fabs(ss[0]) <= epsilon) ss[0] = 0.0; if (fabs(ss[1]) <= epsilon) ss[1] = 0.0; if (fabs(ss[2]) <= epsilon) ss[2] = 0.0; if (fabs(ss[3]) <= epsilon) ss[3] = 0.0; if (fabs(ss[4]) <= epsilon) ss[4] = 0.0; if (fabs(ss[5]) <= epsilon) ss[5] = 0.0; if (fabs(ss[0]) >= fabs(ss[2])) { j = 0; if (fabs(ss[0]) < fabs(ss[5])) j = 2; } else if (fabs(ss[2]) >= fabs(ss[5])) j = 1; else j = 2; d = 0.0; j = 3 * j; for (i = 0; i<3; i++) { k = ip[i + j]; a[i][l] = ss[k]; d = d + ss[k] * ss[k]; } //if( d > 0.0 ) d = 1.0 / sqrt(d); if (d > epsilon) d = 1.0 / sqrt(d); else d = 0.0; for (i = 0; i<3; i++) a[i][l] = a[i][l] * d; }//for l d = a[0][0] * a[0][2] + a[1][0] * a[1][2] + a[2][0] * a[2][2]; if ((e[0] - e[1]) >(e[1] - e[2])) { m1 = 2; m = 0; } else { m1 = 0; m = 2; } p = 0; for (i = 0; i<3; i++) { a[i][m1] = a[i][m1] - d*a[i][m]; p = p + a[i][m1] * a[i][m1]; } if (p <= tol) { p = 1.0; for (i = 0; i<3; i++) { if (p < fabs(a[i][m])) continue; p = fabs(a[i][m]); j = i; } k = ip2312[j]; l = ip2312[j + 1]; p = sqrt(a[k][m] * a[k][m] + a[l][m] * a[l][m]); if (p > tol) { a[j][m1] = 0.0; a[k][m1] = -a[l][m] / p; a[l][m1] = a[k][m] / p; } else a_failed = 1; }//if p<=tol else { p = 1.0 / sqrt(p); for (i = 0; i<3; i++) a[i][m1] = a[i][m1] * p; }//else p<=tol if (a_failed != 1) { a[0][1] = a[1][2] * a[2][0] - a[1][0] * a[2][2]; a[1][1] = a[2][2] * a[0][0] - a[2][0] * a[0][2]; a[2][1] = a[0][2] * a[1][0] - a[0][0] * a[1][2]; } }//if(mode!=0) }//h>0 //compute b anyway if (mode != 0 && a_failed != 1)//a is computed correctly { //compute b for (l = 0; l<2; l++) { d = 0.0; for (i = 0; i<3; i++) { b[i][l] = r[i][0] * a[0][l] + r[i][1] * a[1][l] + r[i][2] * a[2][l]; d = d + b[i][l] * b[i][l]; } //if( d > 0 ) d = 1.0 / sqrt(d); if (d > epsilon) d = 1.0 / sqrt(d); else d = 0.0; for (i = 0; i<3; i++) b[i][l] = b[i][l] * d; } d = b[0][0] * b[0][1] + b[1][0] * b[1][1] + b[2][0] * b[2][1]; p = 0.0; for (i = 0; i<3; i++) { b[i][1] = b[i][1] - d*b[i][0]; p += b[i][1] * b[i][1]; } if (p <= tol) { p = 1.0; for (i = 0; i<3; i++) { if (p tol) { b[j][1] = 0.0; b[k][1] = -b[l][0] / p; b[l][1] = b[k][0] / p; } else b_failed = 1; }//if( p <= tol ) else { p = 1.0 / sqrt(p); for (i = 0; i<3; i++) b[i][1] = b[i][1] * p; } if (b_failed != 1) { b[0][2] = b[1][0] * b[2][1] - b[1][1] * b[2][0]; b[1][2] = b[2][0] * b[0][1] - b[2][1] * b[0][0]; b[2][2] = b[0][0] * b[1][1] - b[0][1] * b[1][0]; //compute u for (i = 0; i<3; i++) for (j = 0; j<3; j++) u[i][j] = b[i][0] * a[j][0] + b[i][1] * a[j][1] + b[i][2] * a[j][2]; } //compute t for (i = 0; i<3; i++) t[i] = ((yc[i] - u[i][0] * xc[0]) - u[i][1] * xc[1]) - u[i][2] * xc[2]; }//if(mode!=0 && a_failed!=1) }//spur>0 else //just compute t and errors { //compute t for (i = 0; i<3; i++) t[i] = ((yc[i] - u[i][0] * xc[0]) - u[i][1] * xc[1]) - u[i][2] * xc[2]; }//else spur>0 //compute rms for (i = 0; i<3; i++) { if (e[i] < 0) e[i] = 0; e[i] = sqrt(e[i]); } d = e[2]; if (sigma < 0.0) d = -d; d = (d + e[1]) + e[0]; if (mode == 2 || mode == 0) { rms1 = (e0 - d) - d; if (rms1 < 0.0) rms1 = 0.0; } *rms = rms1; return true; } /* TMalign.h" */ /* Functions for the core TMalign algorithm, including the entry function * TMalign_main */ // 1, collect those residues with dis3) { inc++; double dinc=(d+inc*0.5); d_tmp = dinc * dinc; } else break; } *score1=score_sum/Lnorm; return n_cut; } int score_fun8_standard(double **xa, double **ya, int n_ali, double d, int i_ali[], double *score1, int score_sum_method, double score_d8, double d0) { double score_sum = 0, di; double d_tmp = d*d; double d02 = d0*d0; double score_d8_cut = score_d8*score_d8; int i, n_cut, inc = 0; while (1) { n_cut = 0; score_sum = 0; for (i = 0; i3) { inc++; double dinc = (d + inc*0.5); d_tmp = dinc * dinc; } else break; } *score1 = score_sum / n_ali; return n_cut; } double TMscore8_search(double **r1, double **r2, double **xtm, double **ytm, double **xt, int Lali, double t0[3], double u0[3][3], int simplify_step, int score_sum_method, double *Rcomm, double local_d0_search, double Lnorm, double score_d8, double d0) { int i, m; double score_max, score, rmsd; const int kmax=Lali; int k_ali[kmax], ka, k; double t[3]; double u[3][3]; double d; //iterative parameters int n_it=20; //maximum number of iterations int n_init_max=6; //maximum number of different fragment length int L_ini[n_init_max]; //fragment lengths, Lali, Lali/2, Lali/4 ... 4 int L_ini_min=4; if(Laliscore_max) { score_max=score; //save the rotation matrix for(k=0; k<3; k++) { t0[k]=t[k]; u0[k][0]=u[k][0]; u0[k][1]=u[k][1]; u0[k][2]=u[k][2]; } } //try to extend the alignment iteratively d = local_d0_search + 1; for(int it=0; itscore_max) { score_max=score; //save the rotation matrix for(k=0; k<3; k++) { t0[k]=t[k]; u0[k][0]=u[k][0]; u0[k][1]=u[k][1]; u0[k][2]=u[k][2]; } } //check if it converges if(n_cut==ka) { for(k=0; kiL_max) i=iL_max; //do this to use the last missed fragment } else if(i>=iL_max) break; }//while(1) //end of one fragment }//for(i_init return score_max; } double TMscore8_search_standard( double **r1, double **r2, double **xtm, double **ytm, double **xt, int Lali, double t0[3], double u0[3][3], int simplify_step, int score_sum_method, double *Rcomm, double local_d0_search, double score_d8, double d0) { int i, m; double score_max, score, rmsd; const int kmax = Lali; int k_ali[kmax], ka, k; double t[3]; double u[3][3]; double d; //iterative parameters int n_it = 20; //maximum number of iterations int n_init_max = 6; //maximum number of different fragment length int L_ini[n_init_max]; //fragment lengths, Lali, Lali/2, Lali/4 ... 4 int L_ini_min = 4; if (Laliscore_max) { score_max = score; //save the rotation matrix for (k = 0; k<3; k++) { t0[k] = t[k]; u0[k][0] = u[k][0]; u0[k][1] = u[k][1]; u0[k][2] = u[k][2]; } } //try to extend the alignment iteratively d = local_d0_search + 1; for (int it = 0; itscore_max) { score_max = score; //save the rotation matrix for (k = 0; k<3; k++) { t0[k] = t[k]; u0[k][0] = u[k][0]; u0[k][1] = u[k][1]; u0[k][2] = u[k][2]; } } //check if it converges if (n_cut == ka) { for (k = 0; kiL_max) i = iL_max; //do this to use the last missed fragment } else if (i >= iL_max) break; }//while(1) //end of one fragment }//for(i_init return score_max; } //Comprehensive TMscore search engine // input: two vector sets: x, y // an alignment invmap0[] between x and y // simplify_step: 1 or 40 or other integers // score_sum_method: 0 for score over all pairs // 8 for socre over the pairs with dist=0) //aligned { xtm[k][0]=x[j][0]; xtm[k][1]=x[j][1]; xtm[k][2]=x[j][2]; ytm[k][0]=y[i][0]; ytm[k][1]=y[i][1]; ytm[k][2]=y[i][2]; k++; } } //detailed search 40-->1 tmscore = TMscore8_search(r1, r2, xtm, ytm, xt, k, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); return tmscore; } double detailed_search_standard( double **r1, double **r2, double **xtm, double **ytm, double **xt, double **x, double **y, int xlen, int ylen, int invmap0[], double t[3], double u[3][3], int simplify_step, int score_sum_method, double local_d0_search, const bool& bNormalize, double Lnorm, double score_d8, double d0) { //x is model, y is template, try to superpose onto y int i, j, k; double tmscore; double rmsd; k=0; for(i=0; i=0) //aligned { xtm[k][0]=x[j][0]; xtm[k][1]=x[j][1]; xtm[k][2]=x[j][2]; ytm[k][0]=y[i][0]; ytm[k][1]=y[i][1]; ytm[k][2]=y[i][2]; k++; } } //detailed search 40-->1 tmscore = TMscore8_search_standard( r1, r2, xtm, ytm, xt, k, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, score_d8, d0); if (bNormalize)// "-i", to use standard_TMscore, then bNormalize=true, else bNormalize=false; tmscore = tmscore * k / Lnorm; return tmscore; } //compute the score quickly in three iterations double get_score_fast( double **r1, double **r2, double **xtm, double **ytm, double **x, double **y, int xlen, int ylen, int invmap[], double d0, double d0_search, double t[3], double u[3][3]) { double rms, tmscore, tmscore1, tmscore2; int i, j, k; k=0; for(j=0; j=0) { r1[k][0]=x[i][0]; r1[k][1]=x[i][1]; r1[k][2]=x[i][2]; r2[k][0]=y[j][0]; r2[k][1]=y[j][1]; r2[k][2]=y[j][2]; xtm[k][0]=x[i][0]; xtm[k][1]=x[i][1]; xtm[k][2]=x[i][2]; ytm[k][0]=y[j][0]; ytm[k][1]=y[j][1]; ytm[k][2]=y[j][2]; k++; } else if(i!=-1) PrintErrorAndQuit("Wrong map!\n"); } Kabsch(r1, r2, k, 1, &rms, t, u); //evaluate score double di; const int len=k; double dis[len]; double d00=d0_search; double d002=d00*d00; double d02=d0*d0; int n_ali=k; double xrot[3]; tmscore=0; for(k=0; k dis_vec(dis, dis+n_ali); sort(dis_vec.begin(), dis_vec.end()); if (d002t3) d002t += 0.5; else break; } if(n_ali!=j) { Kabsch(r1, r2, j, 1, &rms, t, u); tmscore1=0; for(k=0; k dis_vec(dis, dis+n_ali); sort(dis_vec.begin(), dis_vec.end()); if (d002t3) d002t += 0.5; else break; } //evaluate the score Kabsch(r1, r2, j, 1, &rms, t, u); tmscore2=0; for(k=0; k=tmscore) tmscore=tmscore1; if(tmscore2>=tmscore) tmscore=tmscore2; return tmscore; // no need to normalize this score because it will not be used for latter scoring } //perform gapless threading to find the best initial alignment //input: x, y, xlen, ylen //output: y2x0 stores the best alignment: e.g., //y2x0[j]=i means: //the jth element in y is aligned to the ith element in x if i>=0 //the jth element in y is aligned to a gap in x if i==-1 double get_initial(double **r1, double **r2, double **xtm, double **ytm, double **x, double **y, int xlen, int ylen, int *y2x, double d0, double d0_search, const bool fast_opt, double t[3], double u[3][3]) { int min_len=getmin(xlen, ylen); if(min_len<3) PrintErrorAndQuit("Sequence is too short <3!\n"); int min_ali= min_len/2; //minimum size of considered fragment if(min_ali<=5) min_ali=5; int n1, n2; n1 = -ylen+min_ali; n2 = xlen-min_ali; int i, j, k, k_best; double tmscore, tmscore_max=-1; k_best=n1; for(k=n1; k<=n2; k+=(fast_opt)?5:1) { //get the map for(j=0; j=0 && i=tmscore_max) { tmscore_max=tmscore; k_best=k; } } //extract the best map k=k_best; for(j=0; j=0 && icoil, 2->helix, 3->turn, 4->strand */ void make_sec(double **x, int len, char *sec) { int j1, j2, j3, j4, j5; double d13, d14, d15, d24, d25, d35; for(int i=0; i=0 && j5 >&bp, int a, int b,int &c, int &d) { int i; for (i=0;i0) { if (a+iunpair, 2->paired with upstream, 3->paired with downstream */ void make_sec(char *seq, double **x, int len, char *sec,const string atom_opt) { int ii,jj,i,j; float lb=12.5; // lower bound for " C3'" float ub=15.0; // upper bound for " C3'" if (atom_opt==" C4'") {lb=14.0;ub=16.0;} else if(atom_opt==" C5'") {lb=16.0;ub=18.0;} else if(atom_opt==" O3'") {lb=13.5;ub=16.5;} else if(atom_opt==" O5'") {lb=15.5;ub=18.5;} else if(atom_opt==" P ") {lb=16.5;ub=21.0;} float dis; vector bp_tmp(len,false); vector > bp(len,bp_tmp); bp_tmp.clear(); for (i=0; ilb && dis A0_var,B0_var,C0_var,D0_var; for (i=0; i0 && j+1=A0_var[i]&&A0_var[j]<=C0_var[i])|| (C0_var[j]>=A0_var[i]&&C0_var[j]<=C0_var[i])|| (D0_var[j]>=A0_var[i]&&D0_var[j]<=C0_var[i])|| (B0_var[j]>=A0_var[i]&&B0_var[j]<=C0_var[i])|| (A0_var[j]>=D0_var[i]&&A0_var[j]<=B0_var[i])|| (C0_var[j]>=D0_var[i]&&C0_var[j]<=B0_var[i])|| (D0_var[j]>=D0_var[i]&&D0_var[j]<=B0_var[i])|| (B0_var[j]>=D0_var[i]&&B0_var[j]<=B0_var[i])) { sign=-1; break; } } } if(sign!=0) continue; */ for (j=0;;j++) { if(A0_var[i]+j>C0_var[i]) break; sec[A0_var[i]+j]='<'; sec[D0_var[i]+j]='>'; } } sec[len]=0; /* clean up */ A0_var.clear(); B0_var.clear(); C0_var.clear(); D0_var.clear(); bp.clear(); } //get initial alignment from secondary structure alignment //input: x, y, xlen, ylen //output: y2x stores the best alignment: e.g., //y2x[j]=i means: //the jth element in y is aligned to the ith element in x if i>=0 //the jth element in y is aligned to a gap in x if i==-1 void get_initial_ss(bool **path, double **val, const char *secx, const char *secy, int xlen, int ylen, int *y2x) { double gap_open=-1.0; NWDP_TM(path, val, secx, secy, xlen, ylen, gap_open, y2x); } // get_initial5 in TMalign fortran, get_initial_local in TMalign c by yangji //get initial alignment of local structure superposition //input: x, y, xlen, ylen //output: y2x stores the best alignment: e.g., //y2x[j]=i means: //the jth element in y is aligned to the ith element in x if i>=0 //the jth element in y is aligned to a gap in x if i==-1 bool get_initial5( double **r1, double **r2, double **xtm, double **ytm, bool **path, double **val, double **x, double **y, int xlen, int ylen, int *y2x, double d0, double d0_search, const bool fast_opt, const double D0_MIN) { double GL, rmsd; double t[3]; double u[3][3]; double d01 = d0 + 1.5; if (d01 < D0_MIN) d01 = D0_MIN; double d02 = d01*d01; double GLmax = 0; int aL = getmin(xlen, ylen); int *invmap = new int[ylen + 1]; // jump on sequence1--------------> int n_jump1 = 0; if (xlen > 250) n_jump1 = 45; else if (xlen > 200) n_jump1 = 35; else if (xlen > 150) n_jump1 = 25; else n_jump1 = 15; if (n_jump1 > (xlen / 3)) n_jump1 = xlen / 3; // jump on sequence2--------------> int n_jump2 = 0; if (ylen > 250) n_jump2 = 45; else if (ylen > 200) n_jump2 = 35; else if (ylen > 150) n_jump2 = 25; else n_jump2 = 15; if (n_jump2 > (ylen / 3)) n_jump2 = ylen / 3; // fragment to superimpose--------------> int n_frag[2] = { 20, 100 }; if (n_frag[0] > (aL / 3)) n_frag[0] = aL / 3; if (n_frag[1] > (aL / 2)) n_frag[1] = aL / 2; // start superimpose search--------------> if (fast_opt) { n_jump1*=5; n_jump2*=5; } bool flag = false; for (int i_frag = 0; i_frag < 2; i_frag++) { int m1 = xlen - n_frag[i_frag] + 1; int m2 = ylen - n_frag[i_frag] + 1; for (int i = 0; iGLmax) { GLmax = GL; for (int ii = 0; ii=0) { r1[k][0]=x[i][0]; r1[k][1]=x[i][1]; r1[k][2]=x[i][2]; r2[k][0]=y[j][0]; r2[k][1]=y[j][1]; r2[k][2]=y[j][2]; k++; } } Kabsch(r1, r2, k, 1, &rmsd, t, u); for(int ii=0; ii=0 //the jth element in y is aligned to a gap in x if i==-1 void get_initial_ssplus(double **r1, double **r2, double **score, bool **path, double **val, const char *secx, const char *secy, double **x, double **y, int xlen, int ylen, int *y2x0, int *y2x, const double D0_MIN, double d0) { //create score matrix for DP score_matrix_rmsd_sec(r1, r2, score, secx, secy, x, y, xlen, ylen, y2x0, D0_MIN,d0); double gap_open=-1.0; NWDP_TM(score, path, val, xlen, ylen, gap_open, y2x); } void find_max_frag(double **x, int len, int *start_max, int *end_max, double dcu0, const bool fast_opt) { int r_min, fra_min=4; //minimum fragment for search if (fast_opt) fra_min=8; int start; int Lfr_max=0; r_min= (int) (len*1.0/3.0); //minimum fragment, in case too small protein if(r_min > fra_min) r_min=fra_min; int inc=0; double dcu0_cut=dcu0*dcu0;; double dcu_cut=dcu0_cut; while(Lfr_max < r_min) { Lfr_max=0; int j=1; //number of residues at nf-fragment start=0; for(int i=1; i Lfr_max) { Lfr_max=j; *start_max=start; *end_max=i; } j=1; } } else { if(j>Lfr_max) { Lfr_max=j; *start_max=start; *end_max=i-1; } j=1; start=i; } }// for i; if(Lfr_max < r_min) { inc++; double dinc=pow(1.1, (double) inc) * dcu0; dcu_cut= dinc*dinc; } }//while <; } //perform fragment gapless threading to find the best initial alignment //input: x, y, xlen, ylen //output: y2x0 stores the best alignment: e.g., //y2x0[j]=i means: //the jth element in y is aligned to the ith element in x if i>=0 //the jth element in y is aligned to a gap in x if i==-1 double get_initial_fgt(double **r1, double **r2, double **xtm, double **ytm, double **x, double **y, int xlen, int ylen, int *y2x, double d0, double d0_search, double dcu0, const bool fast_opt, double t[3], double u[3][3]) { int fra_min=4; //minimum fragment for search if (fast_opt) fra_min=8; int fra_min1=fra_min-1; //cutoff for shift, save time int xstart=0, ystart=0, xend=0, yend=0; find_max_frag(x, xlen, &xstart, &xend, dcu0, fast_opt); find_max_frag(y, ylen, &ystart, ¥d, dcu0, fast_opt); int Lx = xend-xstart+1; int Ly = yend-ystart+1; int *ifr, *y2x_; int L_fr=getmin(Lx, Ly); ifr= new int[L_fr]; y2x_= new int[ylen+1]; //select what piece will be used. The original implement may cause //asymetry, but only when xlen==ylen and Lx==Ly //if L1=Lfr1 and L2=Lfr2 (normal proteins), it will be the same as initial1 if(LxLy || (Lx==Ly && xlen>ylen)) { for(int i=0; i=0 && i=tmscore_max) { tmscore_max=tmscore; for(j=0; j=0 && i=tmscore_max) { tmscore_max=tmscore; for(j=0; j=0 && i=tmscore_max) { tmscore_max=tmscore; for(j=0; j=0 && i=tmscore_max) { tmscore_max=tmscore; for(j=0; j=0) //aligned { xtm[k][0]=x[i][0]; xtm[k][1]=x[i][1]; xtm[k][2]=x[i][2]; ytm[k][0]=y[j][0]; ytm[k][1]=y[j][1]; ytm[k][2]=y[j][2]; k++; } } tmscore = TMscore8_search(r1, r2, xtm, ytm, xt, k, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); if(tmscore>tmscore_max) { tmscore_max=tmscore; for(i=0; i0) { if(fabs(tmscore_old-tmscore)<0.000001) break; } tmscore_old=tmscore; }// for iteration }//for gapopen delete []invmap; return tmscore_max; } void output_pymol(const string xname, const string yname, const string fname_super, double t[3], double u[3][3], const int ter_opt, const int mm_opt, const int split_opt, const int mirror_opt, const char *seqM, const char *seqxA, const char *seqyA, const vector&resi_vec1, const vector&resi_vec2, const string chainID1, const string chainID2) { int compress_type=0; // uncompressed file ifstream fin; #ifndef REDI_PSTREAM_H_SEEN ifstream fin_gz; #else redi::ipstream fin_gz; // if file is compressed if (xname.size()>=3 && xname.substr(xname.size()-3,3)==".gz") { fin_gz.open("gunzip -c "+xname); compress_type=1; } else if (xname.size()>=4 && xname.substr(xname.size()-4,4)==".bz2") { fin_gz.open("bzcat "+xname); compress_type=2; } else #endif fin.open(xname.c_str()); stringstream buf; stringstream buf_pymol; string line; double x[3]; // before transform double x1[3]; // after transform /* for PDBx/mmCIF only */ map _atom_site; size_t atom_site_pos; vector line_vec; int infmt=-1; // 0 - PDB, 3 - PDBx/mmCIF while (compress_type?fin_gz.good():fin.good()) { if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0) // PDB format { infmt=0; x[0]=atof(line.substr(30,8).c_str()); x[1]=atof(line.substr(38,8).c_str()); x[2]=atof(line.substr(46,8).c_str()); if (mirror_opt) x[2]=-x[2]; transform(t, u, x, x1); buf<=1 && line.compare(0,3,"END")==0) break; } } if (compress_type) fin_gz.close(); else fin.close(); string fname_super_full=fname_super; if (infmt==0) fname_super_full+=".pdb"; else if (infmt==3) fname_super_full+=".cif"; ofstream fp; fp.open(fname_super_full.c_str()); fp<=1) // align one chain from model 1 { chain1_sele=" and c. "+chainID1.substr(1); chain2_sele=" and c. "+chainID2.substr(1); } else if (split_opt==2 && ter_opt==0) // align one chain from each model { for (i=1;i pml_list; pml_list.push_back(fname_super+""); pml_list.push_back(fname_super+"_atm"); pml_list.push_back(fname_super+"_all"); pml_list.push_back(fname_super+"_all_atm"); pml_list.push_back(fname_super+"_all_atm_lig"); for (int p=0;p&resi_vec1, const vector&resi_vec2, const string chainID1, const string chainID2, const int xlen, const int ylen, const double d0A, const int n_ali8, const double rmsd, const double TM1, const double Liden) { stringstream buf; stringstream buf_all; stringstream buf_atm; stringstream buf_all_atm; stringstream buf_all_atm_lig; //stringstream buf_pdb; stringstream buf_tm; string line; double x[3]; // before transform double x1[3]; // after transform bool after_ter; // true if passed the "TER" line in PDB string asym_id; // chain ID buf_tm<<"REMARK US-align" <<"\nREMARK Structure 1:"<=1) // align one chain from model 1 { chain1_sele=chainID1.substr(1); chain2_sele=chainID2.substr(1); } else if (split_opt==2 && ter_opt==0) // align one chain from each model { for (i=1;i _atom_site; int atom_site_pos; vector line_vec; string atom; // 4-character atom name string AA; // 3-character residue name string resi; // 4-character residue sequence number string inscode; // 1-character insertion code string model_index; // model index bool is_mmcif=false; /* used for CONECT record of chain1 */ int ca_idx1=0; // all CA atoms int lig_idx1=0; // all atoms vector idx_vec; /* used for CONECT record of chain2 */ int ca_idx2=0; // all CA atoms int lig_idx2=0; // all atoms /* extract aligned region */ vector resi_aln1; vector resi_aln2; int i1=-1; int i2=-1; if (!mm_opt) { for (i=0;i=3 && line.compare(0,3,"TER")==0) after_ter=true; if (is_mmcif==false && line.size()>=54 && (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0)) // PDB format { if (line[16]!='A' && line[16]!=' ') continue; x[0]=atof(line.substr(30,8).c_str()); x[1]=atof(line.substr(38,8).c_str()); x[2]=atof(line.substr(46,8).c_str()); if (mirror_opt) x[2]=-x[2]; transform(t, u, x, x1); //buf_pdb<=2) { if (ca_idx1 && asym_id.size() && asym_id!=line.substr(21,1)) { after_ter=true; continue; } asym_id=line[21]; } buf_all_atm<<"ATOM "<=5) atom=atom.substr(0,4); AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) AA=AA.substr(0,3); if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; while (resi.size()<4) resi=' '+resi; if (resi.size()>4) resi=resi.substr(0,4); inscode=' '; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") inscode=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; if (_atom_site.count("auth_asym_id")) { if (chain1_sele.size()) after_ter =line_vec[_atom_site["auth_asym_id"]]!=chain1_sele; else if (ter_opt>=2 && ca_idx1 && asym_id.size() && asym_id!=line_vec[_atom_site["auth_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["auth_asym_id"]]; } else if (_atom_site.count("label_asym_id")) { if (chain1_sele.size()) after_ter =line_vec[_atom_site["label_asym_id"]]!=chain1_sele; if (ter_opt>=2 && ca_idx1 && asym_id.size() && asym_id!=line_vec[_atom_site["label_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["label_asym_id"]]; } //buf_pdb<=1 && line.compare(0,3,"END")==0) break; } } fin.close(); if (!mm_opt) buf<<"TER\n"; buf_all<<"TER\n"; if (!mm_opt) buf_atm<<"TER\n"; buf_all_atm<<"TER\n"; buf_all_atm_lig<<"TER\n"; for (i=1;i=3 && line.compare(0,3,"TER")==0) after_ter=true; if (line.size()>=54 && (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0)) // PDB format { if (line[16]!='A' && line[16]!=' ') continue; if (after_ter && line.compare(0,6,"ATOM ")==0) continue; lig_idx2++; buf_all_atm_lig<=2) { if (ca_idx2 && asym_id.size() && asym_id!=line.substr(21,1)) { after_ter=true; continue; } asym_id=line[21]; } buf_all_atm<<"ATOM "<=5) atom=atom.substr(0,4); AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) AA=AA.substr(0,3); if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; while (resi.size()<4) resi=' '+resi; if (resi.size()>4) resi=resi.substr(0,4); inscode=' '; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") inscode=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; if (_atom_site.count("auth_asym_id")) { if (chain2_sele.size()) after_ter =line_vec[_atom_site["auth_asym_id"]]!=chain2_sele; if (ter_opt>=2 && ca_idx2 && asym_id.size() && asym_id!=line_vec[_atom_site["auth_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["auth_asym_id"]]; } else if (_atom_site.count("label_asym_id")) { if (chain2_sele.size()) after_ter =line_vec[_atom_site["label_asym_id"]]!=chain2_sele; if (ter_opt>=2 && ca_idx2 && asym_id.size() && asym_id!=line_vec[_atom_site["label_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["label_asym_id"]]; } if (after_ter==false || line_vec[_atom_site["group_PDB"]]=="HETATM") { lig_idx2++; buf_all_atm_lig<=1 && line.compare(0,3,"END")==0) break; } } fin.close(); if (!mm_opt) buf<<"TER\n"; buf_all<<"TER\n"; if (!mm_opt) buf_atm<<"TER\n"; buf_all_atm<<"TER\n"; buf_all_atm_lig<<"TER\n"; for (i=ca_idx1+1;i pml_list; pml_list.push_back(fname_super+""); pml_list.push_back(fname_super+"_atm"); pml_list.push_back(fname_super+"_all"); pml_list.push_back(fname_super+"_all_atm"); pml_list.push_back(fname_super+"_all_atm_lig"); for (i=0;i&resi_vec1, const vector&resi_vec2) { if (outfmt_opt<=0) { printf("\nName of Structure_1: %s%s (to be superimposed onto Structure_2)\n", xname.c_str(), chainID1.c_str()); printf("Name of Structure_2: %s%s\n", yname.c_str(), chainID2.c_str()); printf("Length of Structure_1: %d residues\n", xlen); printf("Length of Structure_2: %d residues\n\n", ylen); if (i_opt) printf("User-specified initial alignment: TM/Lali/rmsd = %7.5lf, %4d, %6.3lf\n", TM_ali, L_ali, rmsd_ali); printf("Aligned length= %d, RMSD= %6.2f, Seq_ID=n_identical/n_aligned= %4.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); printf("TM-score= %6.5f (normalized by length of Structure_1: L=%d, d0=%.2f)\n", TM2, xlen, d0B); printf("TM-score= %6.5f (normalized by length of Structure_2: L=%d, d0=%.2f)\n", TM1, ylen, d0A); if (a_opt==1) printf("TM-score= %6.5f (if normalized by average length of two structures: L=%.1f, d0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if (u_opt) printf("TM-score= %6.5f (normalized by user-specified L=%.2f and d0=%.2f)\n", TM4, Lnorm_ass, d0u); if (d_opt) printf("TM-score= %6.5f (scaled by user-specified d0=%.2f, and L=%d)\n", TM5, d0_scale, ylen); printf("(You should use TM-score normalized by length of the reference structure)\n"); //output alignment printf("\n(\":\" denotes residue pairs of d <%4.1f Angstrom, ", d0_out); printf("\".\" denotes other aligned residues)\n"); printf("%s\n", seqxA); printf("%s\n", seqM); printf("%s\n", seqyA); } else if (outfmt_opt==1) { printf(">%s%s\tL=%d\td0=%.2f\tseqID=%.3f\tTM-score=%.5f\n", xname.c_str(), chainID1.c_str(), xlen, d0B, Liden/xlen, TM2); printf("%s\n", seqxA); printf(">%s%s\tL=%d\td0=%.2f\tseqID=%.3f\tTM-score=%.5f\n", yname.c_str(), chainID2.c_str(), ylen, d0A, Liden/ylen, TM1); printf("%s\n", seqyA); printf("# Lali=%d\tRMSD=%.2f\tseqID_ali=%.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); if (i_opt) printf("# User-specified initial alignment: TM=%.5lf\tLali=%4d\trmsd=%.3lf\n", TM_ali, L_ali, rmsd_ali); if(a_opt) printf("# TM-score=%.5f (normalized by average length of two structures: L=%.1f\td0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if(u_opt) printf("# TM-score=%.5f (normalized by user-specified L=%.2f\td0=%.2f)\n", TM4, Lnorm_ass, d0u); if(d_opt) printf("# TM-score=%.5f (scaled by user-specified d0=%.2f\tL=%d)\n", TM5, d0_scale, ylen); printf("$$$$\n"); } else if (outfmt_opt==2) { printf("%s%s\t%s%s\t%.4f\t%.4f\t%.2f\t%4.3f\t%4.3f\t%4.3f\t%d\t%d\t%d", xname.c_str(), chainID1.c_str(), yname.c_str(), chainID2.c_str(), TM2, TM1, rmsd, Liden/xlen, Liden/ylen, (n_ali8>0)?Liden/n_ali8:0, xlen, ylen, n_ali8); } if (outfmt_opt<5) cout << endl; if (strlen(fname_matrix)) output_rotation_matrix(fname_matrix, t, u); if (o_opt==1) output_pymol(xname, yname, fname_super, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2); else if (o_opt==2) output_rasmol(xname, yname, fname_super, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2, xlen, ylen, d0A, n_ali8, rmsd, TM1, Liden); } void output_mTMalign_results(const string xname, const string yname, const string chainID1, const string chainID2, const int xlen, const int ylen, double t[3], double u[3][3], const double TM1, const double TM2, const double TM3, const double TM4, const double TM5, const double rmsd, const double d0_out, const char *seqM, const char *seqxA, const char *seqyA, const double Liden, const int n_ali8, const int L_ali, const double TM_ali, const double rmsd_ali, const double TM_0, const double d0_0, const double d0A, const double d0B, const double Lnorm_ass, const double d0_scale, const double d0a, const double d0u, const char* fname_matrix, const int outfmt_opt, const int ter_opt, const int mm_opt, const int split_opt, const int o_opt, const string fname_super, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const int mirror_opt, const vector&resi_vec1, const vector&resi_vec2) { if (outfmt_opt<=0) { printf("Average aligned length= %d, RMSD= %6.2f, Seq_ID=n_identical/n_aligned= %4.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); printf("Average TM-score= %6.5f (normalized by length of shorter structure: L=%d, d0=%.2f)\n", TM2, xlen, d0B); printf("Average TM-score= %6.5f (normalized by length of longer structure: L=%d, d0=%.2f)\n", TM1, ylen, d0A); if (a_opt==1) printf("Average TM-score= %6.5f (if normalized by average length of two structures: L=%.1f, d0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if (u_opt) printf("Average TM-score= %6.5f (normalized by average L=%.2f and d0=%.2f)\n", TM4, Lnorm_ass, d0u); if (d_opt) printf("Average TM-score= %6.5f (scaled by user-specified d0=%.2f, and L=%d)\n", TM5, d0_scale, ylen); //output alignment printf("In the following, seqID=n_identical/L.\n\n%s\n", seqM); } else if (outfmt_opt==1) { printf("%s\n", seqM); printf("# Lali=%d\tRMSD=%.2f\tseqID_ali=%.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); if (i_opt) printf("# User-specified initial alignment: TM=%.5lf\tLali=%4d\trmsd=%.3lf\n", TM_ali, L_ali, rmsd_ali); if(a_opt) printf("# TM-score=%.5f (normalized by average length of two structures: L=%.1f\td0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if(u_opt) printf("# TM-score=%.5f (normalized by average L=%.2f\td0=%.2f)\n", TM4, Lnorm_ass, d0u); if(d_opt) printf("# TM-score=%.5f (scaled by user-specified d0=%.2f\tL=%d)\n", TM5, d0_scale, ylen); printf("$$$$\n"); } else if (outfmt_opt==2) { printf("%s%s\t%s%s\t%.4f\t%.4f\t%.2f\t%4.3f\t%4.3f\t%4.3f\t%d\t%d\t%d", xname.c_str(), chainID1.c_str(), yname.c_str(), chainID2.c_str(), TM2, TM1, rmsd, Liden/xlen, Liden/ylen, (n_ali8>0)?Liden/n_ali8:0, xlen, ylen, n_ali8); } cout << endl; if (strlen(fname_matrix)) output_rotation_matrix(fname_matrix, t, u); if (o_opt==1) output_pymol(xname, yname, fname_super, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2); else if (o_opt==2) output_rasmol(xname, yname, fname_super, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2, xlen, ylen, d0A, n_ali8, rmsd, TM1, Liden); } double standard_TMscore(double **r1, double **r2, double **xtm, double **ytm, double **xt, double **x, double **y, int xlen, int ylen, int invmap[], int& L_ali, double& RMSD, double D0_MIN, double Lnorm, double d0, double d0_search, double score_d8, double t[3], double u[3][3], const int mol_type) { D0_MIN = 0.5; Lnorm = ylen; if (mol_type>0) // RNA { if (Lnorm<=11) d0=0.3; else if(Lnorm>11 && Lnorm<=15) d0=0.4; else if(Lnorm>15 && Lnorm<=19) d0=0.5; else if(Lnorm>19 && Lnorm<=23) d0=0.6; else if(Lnorm>23 && Lnorm<30) d0=0.7; else d0=(0.6*pow((Lnorm*1.0-0.5), 1.0/2)-2.5); } else { if (Lnorm > 21) d0=(1.24*pow((Lnorm*1.0-15), 1.0/3) -1.8); else d0 = D0_MIN; if (d0 < D0_MIN) d0 = D0_MIN; } double d0_input = d0;// Scaled by seq_min double tmscore;// collected alined residues from invmap int n_al = 0; int i; for (int j = 0; j= 0) { xtm[n_al][0] = x[i][0]; xtm[n_al][1] = x[i][1]; xtm[n_al][2] = x[i][2]; ytm[n_al][0] = y[j][0]; ytm[n_al][1] = y[j][1]; ytm[n_al][2] = y[j][2]; r1[n_al][0] = x[i][0]; r1[n_al][1] = x[i][1]; r1[n_al][2] = x[i][2]; r2[n_al][0] = y[j][0]; r2[n_al][1] = y[j][1]; r2[n_al][2] = y[j][2]; n_al++; } else if (i != -1) PrintErrorAndQuit("Wrong map!\n"); } L_ali = n_al; Kabsch(r1, r2, n_al, 0, &RMSD, t, u); RMSD = sqrt( RMSD/(1.0*n_al) ); int temp_simplify_step = 1; int temp_score_sum_method = 0; d0_search = d0_input; double rms = 0.0; tmscore = TMscore8_search_standard(r1, r2, xtm, ytm, xt, n_al, t, u, temp_simplify_step, temp_score_sum_method, &rms, d0_input, score_d8, d0); tmscore = tmscore * n_al / (1.0*Lnorm); return tmscore; } /* copy the value of t and u into t0,u0 */ void copy_t_u(double t[3], double u[3][3], double t0[3], double u0[3][3]) { int i,j; for (i=0;i<3;i++) { t0[i]=t[i]; for (j=0;j<3;j++) u0[i][j]=u[i][j]; } } /* calculate approximate TM-score given rotation matrix */ double approx_TM(const int xlen, const int ylen, const int a_opt, double **xa, double **ya, double t[3], double u[3][3], const int invmap0[], const int mol_type) { double Lnorm_0=ylen; // normalized by the second protein if (a_opt==-2 && xlen>ylen) Lnorm_0=xlen; // longer else if (a_opt==-1 && xlen=0)//aligned { transform(t, u, &xa[i][0], &xtmp[0]); d=sqrt(dist(&xtmp[0], &ya[j][0])); TMtmp+=1/(1+(d/d0)*(d/d0)); //if (d <= score_d8) TMtmp+=1/(1+(d/d0)*(d/d0)); } } TMtmp/=Lnorm_0; return TMtmp; } void clean_up_after_approx_TM(int *invmap0, int *invmap, double **score, bool **path, double **val, double **xtm, double **ytm, double **xt, double **r1, double **r2, const int xlen, const int minlen) { delete [] invmap0; delete [] invmap; DeleteArray(&score, xlen+1); DeleteArray(&path, xlen+1); DeleteArray(&val, xlen+1); DeleteArray(&xtm, minlen); DeleteArray(&ytm, minlen); DeleteArray(&xt, xlen); DeleteArray(&r1, minlen); DeleteArray(&r2, minlen); return; } /* Entry function for TM-align. Return TM-score calculation status: * 0 - full TM-score calculation * 1 - terminated due to exception * 2-7 - pre-terminated due to low TM-score */ int TMalign_main(double **xa, double **ya, const char *seqx, const char *seqy, const char *secx, const char *secy, double t0[3], double u0[3][3], double &TM1, double &TM2, double &TM3, double &TM4, double &TM5, double &d0_0, double &TM_0, double &d0A, double &d0B, double &d0u, double &d0a, double &d0_out, string &seqM, string &seqxA, string &seqyA, vector&do_vec, double &rmsd0, int &L_ali, double &Liden, double &TM_ali, double &rmsd_ali, int &n_ali, int &n_ali8, const int xlen, const int ylen, const vector sequence, const double Lnorm_ass, const double d0_scale, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const bool fast_opt, const int mol_type, const double TMcut=-1) { double D0_MIN; //for d0 double Lnorm; //normalization length double score_d8,d0,d0_search,dcu0;//for TMscore search double t[3], u[3][3]; //Kabsch translation vector and rotation matrix double **score; // Input score table for dynamic programming bool **path; // for dynamic programming double **val; // for dynamic programming double **xtm, **ytm; // for TMscore search engine double **xt; //for saving the superposed version of r_1 or xtm double **r1, **r2; // for Kabsch rotation /***********************/ /* allocate memory */ /***********************/ int minlen = min(xlen, ylen); NewArray(&score, xlen+1, ylen+1); NewArray(&path, xlen+1, ylen+1); NewArray(&val, xlen+1, ylen+1); NewArray(&xtm, minlen, 3); NewArray(&ytm, minlen, 3); NewArray(&xt, xlen, 3); NewArray(&r1, minlen, 3); NewArray(&r2, minlen, 3); /***********************/ /* parameter set */ /***********************/ parameter_set4search(xlen, ylen, D0_MIN, Lnorm, score_d8, d0, d0_search, dcu0); int simplify_step = 40; //for simplified search engine int score_sum_method = 8; //for scoring method, whether only sum over pairs with dis= ylen || i1 >= xlen) kk1 = L; else if (sequence[0][kk1] != '-') invmap[i2] = i1; } } //--------------- 2. Align proteins from original alignment double prevD0_MIN = D0_MIN;// stored for later use int prevLnorm = Lnorm; double prevd0 = d0; TM_ali = standard_TMscore(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, L_ali, rmsd_ali, D0_MIN, Lnorm, d0, d0_search, score_d8, t, u, mol_type); D0_MIN = prevD0_MIN; Lnorm = prevLnorm; d0 = prevd0; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, 40, 8, local_d0_search, true, Lnorm, score_d8, d0); if (TM > TMmax) { TMmax = TM; for (i = 0; iTMmax) TMmax = TM; if (TMcut>0) copy_t_u(t, u, t0, u0); //run dynamic programing iteratively to find the best alignment TM = DP_iter(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.5*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 2; } } /************************************************************/ /* get initial alignment based on secondary structure */ /************************************************************/ get_initial_ss(path, val, secx, secy, xlen, ylen, invmap); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*0.2) { TM = DP_iter(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.52*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 3; } } /************************************************************/ /* get initial alignment based on local superposition */ /************************************************************/ //=initial5 in original TM-align if (get_initial5( r1, r2, xtm, ytm, path, val, xa, ya, xlen, ylen, invmap, d0, d0_search, fast_opt, D0_MIN)) { TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, t, u, invmap, 0, 2, 2, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } } } else cerr << "\n\nWarning: initial alignment from local superposition fail!\n\n" << endl; if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.54*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 4; } } /********************************************************************/ /* get initial alignment by local superposition+secondary structure */ /********************************************************************/ //=initial3 in original TM-align get_initial_ssplus(r1, r2, score, path, val, secx, secy, xa, ya, xlen, ylen, invmap0, invmap, D0_MIN, d0); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.56*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 5; } } /*******************************************************************/ /* get initial alignment based on fragment gapless threading */ /*******************************************************************/ //=initial4 in original TM-align get_initial_fgt(r1, r2, xtm, ytm, xa, ya, xlen, ylen, invmap, d0, d0_search, dcu0, fast_opt, t, u); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, t, u, invmap, 1, 2, 2, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.58*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 6; } } } //************************************************// // get initial alignment from user's input: // //************************************************// if (i_opt>=1 && i_opt<=2)// if input has set parameter for "-i" { for (int j = 0; j < ylen; j++)// Set aligned position to be "-1" invmap[j] = -1; int i1 = -1;// in C version, index starts from zero, not from one int i2 = -1; int L1 = sequence[0].size(); int L2 = sequence[1].size(); int L = min(L1, L2);// Get positions for aligned residues for (int kk1 = 0; kk1 < L; kk1++) { if (sequence[0][kk1] != '-') i1++; if (sequence[1][kk1] != '-') { i2++; if (i2 >= ylen || i1 >= xlen) kk1 = L; else if (sequence[0][kk1] != '-') invmap[i2] = i1; } } //--------------- 2. Align proteins from original alignment double prevD0_MIN = D0_MIN;// stored for later use int prevLnorm = Lnorm; double prevd0 = d0; TM_ali = standard_TMscore(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, L_ali, rmsd_ali, D0_MIN, Lnorm, d0, d0_search, score_d8, t, u, mol_type); D0_MIN = prevD0_MIN; Lnorm = prevLnorm; d0 = prevd0; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, 40, 8, local_d0_search, true, Lnorm, score_d8, d0); if (TM > TMmax) { TMmax = TM; for (i = 0; iTMmax) { TMmax = TM; for (i = 0; i=0) { flag=true; break; } } if(!flag) { cout << "There is no alignment between the two structures! " << "Program stop with no result!" << endl; TM1=TM2=TM3=TM4=TM5=0; return 1; } /* last TM-score pre-termination */ if (TMcut>0) { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.6*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 7; } } //********************************************************************// // Detailed TMscore search engine --> prepare for final TMscore // //********************************************************************// //run detailed TMscore search engine for the best alignment, and //extract the best rotation matrix (t, u) for the best alignment simplify_step=1; if (fast_opt) simplify_step=40; score_sum_method=8; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap0, t, u, simplify_step, score_sum_method, local_d0_search, false, Lnorm, score_d8, d0); //select pairs with dis=0)//aligned { n_ali++; d=sqrt(dist(&xt[i][0], &ya[j][0])); if (d <= score_d8 || (i_opt == 3)) { m1[k]=i; m2[k]=j; xtm[k][0]=xa[i][0]; xtm[k][1]=xa[i][1]; xtm[k][2]=xa[i][2]; ytm[k][0]=ya[j][0]; ytm[k][1]=ya[j][1]; ytm[k][2]=ya[j][2]; r1[k][0] = xt[i][0]; r1[k][1] = xt[i][1]; r1[k][2] = xt[i][2]; r2[k][0] = ya[j][0]; r2[k][1] = ya[j][1]; r2[k][2] = ya[j][2]; k++; } } } n_ali8=k; Kabsch(r1, r2, n_ali8, 0, &rmsd0, t, u);// rmsd0 is used for final output, only recalculate rmsd0, not t & u rmsd0 = sqrt(rmsd0 / n_ali8); //****************************************// // Final TMscore // // Please set parameters for output // //****************************************// double rmsd; simplify_step=1; score_sum_method=0; double Lnorm_0=ylen; //normalized by length of structure A parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0A=d0; d0_0=d0A; local_d0_search = d0_search; TM1 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0 = TM1; //normalized by length of structure B parameter_set4final(xlen+0.0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0B=d0; local_d0_search = d0_search; TM2 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); double Lnorm_d0; if (a_opt>0) { //normalized by average length of structures A, B Lnorm_0=(xlen+ylen)*0.5; parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0a=d0; d0_0=d0a; local_d0_search = d0_search; TM3 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM3; } if (u_opt) { //normalized by user assigned length parameter_set4final(Lnorm_ass, D0_MIN, Lnorm, d0, d0_search, mol_type); d0u=d0; d0_0=d0u; Lnorm_0=Lnorm_ass; local_d0_search = d0_search; TM4 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM4; } if (d_opt) { //scaled by user assigned d0 parameter_set4scale(ylen, d0_scale, Lnorm, d0, d0_search); d0_out=d0_scale; d0_0=d0_scale; //Lnorm_0=ylen; Lnorm_d0=Lnorm_0; local_d0_search = d0_search; TM5 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM5; } /* derive alignment from superposition */ int ali_len=xlen+ylen; //maximum length of alignment seqxA.assign(ali_len,'-'); seqM.assign( ali_len,' '); seqyA.assign(ali_len,'-'); do_vec.clear(); do_vec.assign(ali_len,0); //do_rotation(xa, xt, xlen, t, u); do_rotation(xa, xt, xlen, t0, u0); int kk=0, i_old=0, j_old=0; d=0; Liden=0; //double SO=0; for(int k=0; k &do_vec, double &rmsd0, int &L_ali, double &Liden, double &TM_ali, double &rmsd_ali, int &n_ali, int &n_ali8, const int xlen, const int ylen, const vector sequence, const double Lnorm_ass, const double d0_scale, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const bool fast_opt, const int mol_type, const double TMcut=-1) { char *seqx_cp; // for the protein sequence char *secx_cp; // for the secondary structure double **xa_cp; // coordinates string seqxA_cp,seqyA_cp; // alignment int i,r; int cp_point=0; // position of circular permutation int cp_aln_best=0; // amount of aligned residue in sliding window int cp_aln_current;// amount of aligned residue in sliding window /* duplicate structure */ NewArray(&xa_cp, xlen*2, 3); seqx_cp = new char[xlen*2 + 1]; secx_cp = new char[xlen*2 + 1]; for (r=0;rcp_aln_best) { cp_aln_best=cp_aln_current; cp_point=r; } } seqM.clear(); seqxA.clear(); seqyA.clear(); seqxA_cp.clear(); seqyA_cp.clear(); rmsd0=Liden=n_ali=n_ali8=0; /* fTM-align alignment */ TMalign_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_tmp, d0_scale, 0, false, true, false, true, mol_type, -1); /* do not use circular permutation of number of aligned residues is not * larger than sequence-order dependent alignment */ //cout<<"cp: aln="<=2 should not parse sequence alignment * u_opt corresponds to option -L * if u_opt==2, use d0 from Lnorm_ass for alignment * if hinge>0, append to original invmap */ int se_main( double **xa, double **ya, const char *seqx, const char *seqy, double &TM1, double &TM2, double &TM3, double &TM4, double &TM5, double &d0_0, double &TM_0, double &d0A, double &d0B, double &d0u, double &d0a, double &d0_out, string &seqM, string &seqxA, string &seqyA, vector &do_vec, double &rmsd0, int &L_ali, double &Liden, double &TM_ali, double &rmsd_ali, int &n_ali, int &n_ali8, const int xlen, const int ylen, const vector &sequence, const double Lnorm_ass, const double d0_scale, const bool i_opt, const bool a_opt, const int u_opt, const bool d_opt, const int mol_type, const int outfmt_opt, int *invmap, const int hinge=0) { double D0_MIN; //for d0 double Lnorm; //normalization length double score_d8,d0,d0_search,dcu0;//for TMscore search double **score; // Input score table for dynamic programming bool **path; // for dynamic programming double **val; // for dynamic programming int *m1=NULL; int *m2=NULL; double d; if (outfmt_opt<2) { m1=new int[xlen]; //alignd index in x m2=new int[ylen]; //alignd index in y } /***********************/ /* allocate memory */ /***********************/ NewArray(&score, xlen+1, ylen+1); NewArray(&path, xlen+1, ylen+1); NewArray(&val, xlen+1, ylen+1); int *invmap0 = new int[ylen+1]; int i,j; if (hinge==0) for (j=0;j<=ylen;j++) invmap0[j]=-1; else for (j=0;j seqM_char; if (hinge) { seqM_char.assign(ylen,hinge+'0'); j=-1; for (int r=0;r= ylen || i1 >= xlen) kk1 = L; else if (sequence[0][kk1] != '-') invmap[i2] = i1; } } } if (hinge==0) rmsd0=TM1=TM2=TM3=TM4=TM5=0; else { TM2*=xlen; TM1*=ylen; TM3*=(xlen+ylen)*0.5; TM4*=Lnorm_ass; TM5*=ylen; rmsd0=rmsd0*rmsd0*n_ali8; } int k=0; n_ali=0; n_ali8=0; for(int i=0,j=0; j=0)//aligned { n_ali++; d=sqrt(dist(&xa[i][0], &ya[j][0])); if (d <= score_d8 || i_opt || invmap0[j]==i) { if (outfmt_opt<2) { m1[k]=i; m2[k]=j; } k++; if (invmap0[j]==i) continue; TM2+=1/(1+(d/d0B)*(d/d0B)); // chain_1 TM1+=1/(1+(d/d0A)*(d/d0A)); // chain_2 if (a_opt) TM3+=1/(1+(d/d0a)*(d/d0a)); // -a if (u_opt) TM4+=1/(1+(d/d0u)*(d/d0u)); // -u if (d_opt) TM5+=1/(1+(d/d0_scale)*(d/d0_scale)); // -d rmsd0+=d*d; } else if (hinge) invmap[j]=-1; } } n_ali8=k; TM2/=xlen; TM1/=ylen; TM3/=(xlen+ylen)*0.5; TM4/=Lnorm_ass; TM5/=ylen; if (n_ali8) rmsd0=sqrt(rmsd0/n_ali8); if (outfmt_opt>=2) { if (hinge) seqM_char.clear(); delete []invmap0; DeleteArray(&score, xlen+1); DeleteArray(&path, xlen+1); DeleteArray(&val, xlen+1); return 0; } /* extract aligned sequence */ int ali_len=xlen+ylen; //maximum length of alignment seqxA.assign(ali_len,'-'); seqM.assign( ali_len,' '); seqyA.assign(ali_len,'-'); do_vec.clear(); do_vec.assign(ali_len,0); int kk=0, i_old=0, j_old=0; d=0; Liden=0; for(int k=0; k&mol_vec) { na_chain_num=0; aa_chain_num=0; for (size_t i=0;i0) na_chain_num++; else aa_chain_num++; } return na_chain_num+aa_chain_num; } /* adjust chain assignment for dimer-dimer alignment * return true if assignment is adjusted */ bool adjust_dimer_assignment( const vector > >&xa_vec, const vector > >&ya_vec, const vector&xlen_vec, const vector&ylen_vec, const vector&mol_vec1, const vector&mol_vec2, int *assign1_list, int *assign2_list, const vector >&seqxA_mat, const vector >&seqyA_mat) { /* check currently assigned chains */ int i1,i2,j1,j2; i1=i2=j1=j2=-1; int chain1_num=xa_vec.size(); int i,j; for (i=0;i=0) { if (i1<0) { i1=i; j1=assign1_list[i1]; } else { i2=i; j2=assign1_list[i2]; } } } /* normalize d0 by L */ int xlen=xlen_vec[i1]+xlen_vec[i2]; int ylen=ylen_vec[j1]+ylen_vec[j2]; int mol_type=mol_vec1[i1]+mol_vec1[i2]+ mol_vec2[j1]+mol_vec2[j2]; double D0_MIN, d0, d0_search; double Lnorm=getmin(xlen,ylen); parameter_set4final(getmin(xlen,ylen), D0_MIN, Lnorm, d0, d0_search, mol_type); double **xa,**ya, **xt; NewArray(&xa, xlen, 3); NewArray(&ya, ylen, 3); NewArray(&xt, xlen, 3); double RMSD = 0; double dd = 0; double t[3]; double u[3][3]; size_t L_ali=0; // index of residue in aligned region size_t r=0; // index of residue in full alignment /* total score using current assignment */ L_ali=0; i=j=-1; for (r=0;r=0); return pair_num; } /* assign chain-chain correspondence */ double enhanced_greedy_search(double **TMave_mat,int *assign1_list, int *assign2_list, const int chain1_num, const int chain2_num) { double total_score=0; double tmp_score=0; int i,j; int maxi=0; int maxj=0; /* initialize parameters */ for (i=0;i=0) continue; for (j=0;j=0 || TMave_mat[i][j]<=0) continue; if (TMave_mat[i][j]>tmp_score) { maxi=i; maxj=j; tmp_score=TMave_mat[i][j]; } } } if (tmp_score<=0) break; // error: no assignable chain assign1_list[maxi]=maxj; assign2_list[maxj]=maxi; total_score+=tmp_score; } if (total_score<=0) return total_score; // error: no assignable chain //cout<<"assign1_list={"; //for (i=0;i=0) assign1_tmp[old_i]=old_j; assign2_tmp[j]=i; if (old_j>=0) assign2_tmp[old_j]=old_i; delta_score=TMave_mat[i][j]; if (old_j>=0) delta_score-=TMave_mat[i][old_j]; if (old_i>=0) delta_score-=TMave_mat[old_i][j]; if (old_i>=0 && old_j>=0) delta_score+=TMave_mat[old_i][old_j]; if (delta_score>0) // successful swap { assign1_list[i]=j; if (old_i>=0) assign1_list[old_i]=old_j; assign2_list[j]=i; if (old_j>=0) assign2_list[old_j]=old_i; total_score+=delta_score; break; } else { assign1_tmp[i]=assign1_list[i]; if (old_i>=0) assign1_tmp[old_i]=assign1_list[old_i]; assign2_tmp[j]=assign2_list[j]; if (old_j>=0) assign2_tmp[old_j]=assign2_list[old_j]; } } if (delta_score>0) break; } if (delta_score<=0) break; // cannot swap any chain pair } /* clean up */ delete[]assign1_tmp; delete[]assign2_tmp; return total_score; } double calculate_centroids(const vector > >&a_vec, const int chain_num, double ** centroids) { int L=0; int c,r; // index of chain and residue for (c=0; c d0_vec(chain_num,-1); int c2=0; double d0MM=0; for (c=0; c=3) { /* Kabsch superposition */ Kabsch(r1, r2, Nali, 1, &RMSD, t, u); do_rotation(r1, xt, Nali, t, u); /* calculate pseudo-TMscore */ double dd=0; for (i=0;i=max_TM) max_TM=TMave_mat[i][j]; } } het_deg=(max_TM-min_TM)/max_TM; //cout<<"min_TM="<=0;ut_idx--) { j=ut_tm_vec[ut_idx].second % chain2_num; i=int(ut_tm_vec[ut_idx].second / chain2_num); if (TMave_mat[i][j]<=0) break; if (assign1_tmp[i]>=0 || assign2_tmp[j]>=0) continue; assign1_tmp[i]=j; assign2_tmp[j]=i; TMsum+=TMave_mat[i][j]; TMscore+=ut_tmc_mat[i*chain2_num+j]; //cout<<"ut_idx="<=0) assign1_tmp[old_i]=old_j; assign2_tmp[j]=i; if (old_j>=0) assign2_tmp[old_j]=old_i; MMscore=calMMscore(TMave_mat, assign1_tmp, chain1_num, chain2_num, xcentroids, ycentroids, d0MM, r1, r2, xt, t, u, L); //cout<<"(i,j,old_i,old_j,MMscore)=("<MMscore_old) // successful swap { assign1_list[i]=j; if (old_i>=0) assign1_list[old_i]=old_j; assign2_list[j]=i; if (old_j>=0) assign2_list[old_j]=old_i; delta_score=(MMscore-MMscore_old); MMscore_old=MMscore; //cout<<"MMscore="<=0) assign1_tmp[old_i]=assign1_list[old_i]; assign2_tmp[j]=assign2_list[j]; if (old_j>=0) assign2_tmp[old_j]=assign2_list[old_j]; } } } //cout<<"iter="< >&PDB_lines, const int ter_opt, const int infmt_opt, const int split_opt, const int het_opt) { size_t i=0; // resi i.e. atom index string line; char chainID=0; vector tmp_str_vec; int compress_type=0; // uncompressed file ifstream fin; #ifndef REDI_PSTREAM_H_SEEN ifstream fin_gz; #else redi::ipstream fin_gz; // if file is compressed if (filename.size()>=3 && filename.substr(filename.size()-3,3)==".gz") { fin_gz.open("gunzip -c '"+filename+"'"); compress_type=1; } else if (filename.size()>=4 && filename.substr(filename.size()-4,4)==".bz2") { fin_gz.open("bzcat '"+filename+"'"); compress_type=2; } else #endif fin.open(filename.c_str()); if (infmt_opt==0||infmt_opt==-1) // PDB format { while (compress_type?fin_gz.good():fin.good()) { if (compress_type) getline(fin_gz, line); else getline(fin, line); if (infmt_opt==-1 && line.compare(0,5,"loop_")==0) // PDBx/mmCIF return get_full_PDB_lines(filename,PDB_lines, ter_opt, 3, split_opt,het_opt); if (i > 0) { if (ter_opt>=1 && line.compare(0,3,"END")==0) break; else if (ter_opt>=3 && line.compare(0,3,"TER")==0) break; } if (split_opt && line.compare(0,3,"END")==0) chainID=0; if (line.size()>=54 && (line[16]==' ' || line[16]=='A') && ( (line.compare(0, 6, "ATOM ")==0) || (line.compare(0, 6, "HETATM")==0 && het_opt==1) || (line.compare(0, 6, "HETATM")==0 && het_opt==2 && line.compare(17,3, "MSE")==0))) { if (!chainID) { chainID=line[21]; PDB_lines.push_back(tmp_str_vec); } else if (ter_opt>=2 && chainID!=line[21]) break; if (split_opt==2 && chainID!=line[21]) { chainID=line[21]; PDB_lines.push_back(tmp_str_vec); } PDB_lines.back().push_back(line); i++; } } } else if (infmt_opt==1) // SPICKER format { size_t L=0; float x,y,z; stringstream i8_stream; while (compress_type?fin_gz.good():fin.good()) { if (compress_type) fin_gz>>L>>x>>y>>z; else fin >>L>>x>>y>>z; if (compress_type) getline(fin_gz, line); else getline(fin, line); if (!(compress_type?fin_gz.good():fin.good())) break; for (i=0;i>x>>y>>z; else fin >>x>>y>>z; i8_stream<<"ATOM "< _atom_site; int atom_site_pos; vector line_vec; string alt_id="."; // alternative location indicator string asym_id="."; // this is similar to chainID, except that // chainID is char while asym_id is a string // with possibly multiple char string prev_asym_id=""; string AA=""; // residue name string atom=""; string resi=""; string model_index=""; // the same as model_idx but type is string stringstream i8_stream; while (compress_type?fin_gz.good():fin.good()) { if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.size()==0) continue; if (loop_) loop_ = (line.size()>=2)?(line.compare(0,2,"# ")):(line.compare(0,1,"#")); if (!loop_) { if (line.compare(0,5,"loop_")) continue; while(1) { if (compress_type) { if (fin_gz.good()) getline(fin_gz, line); else PrintErrorAndQuit("ERROR! Unexpected end of "+filename); } else { if (fin.good()) getline(fin, line); else PrintErrorAndQuit("ERROR! Unexpected end of "+filename); } if (line.size()) break; } if (line.compare(0,11,"_atom_site.")) continue; loop_=true; _atom_site.clear(); atom_site_pos=0; _atom_site[Trim(line.substr(11))]=atom_site_pos; while(1) { if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.size()==0) continue; if (line.compare(0,11,"_atom_site.")) break; _atom_site[Trim(line.substr(11))]=++atom_site_pos; } if (_atom_site.count("group_PDB")* _atom_site.count("label_atom_id")* _atom_site.count("label_comp_id")* (_atom_site.count("auth_asym_id")+ _atom_site.count("label_asym_id"))* (_atom_site.count("auth_seq_id")+ _atom_site.count("label_seq_id"))* _atom_site.count("Cartn_x")* _atom_site.count("Cartn_y")* _atom_site.count("Cartn_z")==0) { loop_ = false; cerr<<"Warning! Missing one of the following _atom_site data items: group_PDB, label_atom_id, label_comp_id, auth_asym_id/label_asym_id, auth_seq_id/label_seq_id, Cartn_x, Cartn_y, Cartn_z"<=5) continue; AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) continue; if (_atom_site.count("auth_asym_id")) asym_id=line_vec[_atom_site["auth_asym_id"]]; else asym_id=line_vec[_atom_site["label_asym_id"]]; if (asym_id==".") asym_id=" "; if (_atom_site.count("pdbx_PDB_model_num") && model_index!=line_vec[_atom_site["pdbx_PDB_model_num"]]) { model_index=line_vec[_atom_site["pdbx_PDB_model_num"]]; if (PDB_lines.size() && ter_opt>=1) break; if (PDB_lines.size()==0 || split_opt>=1) { PDB_lines.push_back(tmp_str_vec); prev_asym_id=asym_id; } } if (prev_asym_id!=asym_id) { if (prev_asym_id!="" && ter_opt>=2) break; if (split_opt>=2) PDB_lines.push_back(tmp_str_vec); } if (prev_asym_id!=asym_id) prev_asym_id=asym_id; if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") resi+=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; else resi+=" "; i++; i8_stream<<"ATOM " <&chain_list, const int ter_opt, const int split_opt, const int infmt_opt, const string atom_opt, const int mirror_opt, double **ut_mat, const string&fname_super) { size_t i; int chain_i,a; string name; int chainnum; double x[3]; // before transform double x1[3]; // after transform string line; vector >PDB_lines; int m=0; double t[3]; double u[3][3]; int ui,uj; stringstream buf; string filename; int het_opt=1; for (i=0;i >().swap(PDB_lines); line.clear(); } void parse_chain_list(const vector&chain_list, vector > >&a_vec, vector >&seq_vec, vector >&sec_vec, vector&mol_vec, vector&len_vec, vector&chainID_list, const int ter_opt, const int split_opt, const string mol_opt, const int infmt_opt, const string atom_opt, const bool autojustify, const int mirror_opt, const int het_opt, int &len_aa, int &len_na, const int o_opt, vector&resi_vec, const vector &chain2parse, const vector &model2parse) { size_t i; int chain_i,r; string name; int chainnum; double **xa; int len; char *seq,*sec; vector >PDB_lines; vector tmp_atom_array(3,0); vector > tmp_chain_array; vectortmp_seq_array; vectortmp_sec_array; //vector resi_vec; int read_resi=2; for (i=0;i0 || mol_opt=="RNA") make_sec(seq, xa, len, sec,atom_opt); else make_sec(xa, len, sec); // secondary structure assignment /* store in vector */ tmp_chain_array.assign(len,tmp_atom_array); vectortmp_seq_array(len+1,0); vectortmp_sec_array(len+1,0); for (r=0;r='a' && seq_vec[i][r]<='z') mol_vec[i]++; else mol_vec[i]--; } } } len_aa=0; len_na=0; for (i=0;i0) len_na+=len_vec[i]; else len_aa+=len_vec[i]; } } int copy_chain_pair_data( const vector > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int chain1_num, int chain2_num, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence) { int i,j,r; for (i=0;i > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, double d0_scale, bool fast_opt, const int i_opt=3, const int byresi_opt=0) { double total_score=0; int i,j; int xlen=0; int ylen=0; for (i=0;i do_vec; /* entry function for structure alignment */ TMalign_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, false, true, false, fast_opt, mol_type, -1); /* clean up */ delete [] seqx; delete [] seqy; delete [] secx; delete [] secy; DeleteArray(&xa,xlen); DeleteArray(&ya,ylen); do_vec.clear(); /* re-compute chain level alignment */ for (i=0;i0) Lnorm_ass=len_na; vector sequence_tmp; if (byresi_opt) { sequence_tmp.push_back(seqxA_mat[i][j]); sequence_tmp.push_back(seqyA_mat[i][j]); } /* entry function for structure alignment */ se_main(xt, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence_tmp, Lnorm_ass, d0_scale, byresi_opt, false, 2, false, mol_vec1[i]+mol_vec2[j], 1, invmap); /* print result */ seqxA_mat[i][j]=seqxA; seqyA_mat[i][j]=seqyA; TMave_mat[i][j]=TM4*Lnorm_ass; if (assign1_list[i]==j) total_score+=TMave_mat[i][j]; /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); vector().swap(sequence_tmp); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); delete[]invmap; do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); DeleteArray(&xt,xlen); } if (byresi_opt) { if (sequence.size()<2) sequence.push_back(""); else sequence[0]=""; if (sequence.size()<2) sequence.push_back(""); else sequence[1]=""; for (i=0;i=0) { sequence[0]+=seqxA_mat[i][j]; sequence[1]+=seqyA_mat[i][j]; } } } return total_score; } void MMalign_final( const string xname, const string yname, const vector chainID_list1, const vector chainID_list2, string fname_super, string fname_lign, string fname_matrix, const vector > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqM_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, const double d0_scale, const bool m_opt, const int o_opt, const int outfmt_opt, const int ter_opt, const int split_opt, const bool a_opt, const bool d_opt, const bool fast_opt, const bool full_opt, const int mirror_opt, const vector&resi_vec1, const vector&resi_vec2) { int i,j; int xlen=0; int ylen=0; for (i=0;ido_vec; double Lnorm_ass=len_aa+len_na; /* entry function for structure alignment */ TMalign_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 3, a_opt, false, d_opt, fast_opt, mol_type, -1); /* prepare full complex alignment */ string chainID1=""; string chainID2=""; sequence.clear(); sequence.push_back(""); // seqxA sequence.push_back(""); // seqyA sequence.push_back(""); // seqM int aln_start=0; int aln_end=0; for (i=0;i=0) continue; chainID1+=chainID_list1[i]; chainID2+=':'; string s(seqx_vec[i].begin(),seqx_vec[i].end()); sequence[0]+=s.substr(0,xlen_vec[i])+'*'; sequence[1]+=string(xlen_vec[i],'-')+'*'; s.clear(); sequence[2]+=string(xlen_vec[i],' ')+'*'; } for (j=0;j=0) continue; chainID1+=':'; chainID2+=chainID_list2[j]; string s(seqy_vec[j].begin(),seqy_vec[j].end()); sequence[0]+=string(ylen_vec[j],'-')+'*'; sequence[1]+=s.substr(0,ylen_vec[j])+'*'; s.clear(); sequence[2]+=string(ylen_vec[j],' ')+'*'; } /* print alignment */ output_results(xname, yname, chainID1.c_str(), chainID2.c_str(), xlen, ylen, t0, u0, TM1, TM2, TM3, TM4, TM5, rmsd0, d0_out, sequence[2].c_str(), sequence[0].c_str(), sequence[1].c_str(), Liden, n_ali8, L_ali, TM_ali, rmsd_ali, TM_0, d0_0, d0A, d0B, 0, d0_scale, d0a, d0u, (m_opt?fname_matrix:"").c_str(), outfmt_opt, ter_opt, true, split_opt, o_opt, fname_super, false, a_opt, false, d_opt, mirror_opt, resi_vec1, resi_vec2); /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); delete [] seqx; delete [] seqy; delete [] secx; delete [] secy; DeleteArray(&xa,xlen); DeleteArray(&ya,ylen); sequence[0].clear(); sequence[1].clear(); sequence[2].clear(); do_vec.clear(); if (!full_opt) return; if (outfmt_opt<=2) cout<<"# End of alignment for full complex. The following blocks list alignments for individual chains."<0) Lnorm_ass=len_na; sequence[0]=seqxA_mat[i][j]; sequence[1]=seqyA_mat[i][j]; /* entry function for structure alignment */ se_main(xt, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 1, a_opt, 2, d_opt, mol_vec1[i]+mol_vec2[j], 1, invmap); //TM2=TM4*Lnorm_ass/xlen; //TM1=TM4*Lnorm_ass/ylen; //d0A=d0u; //d0B=d0u; TMave_mat[i][j]=TM4*Lnorm_ass; /* print result */ if (j==assign1_list[i]) output_results(xname, yname, chainID_list1[i].c_str(), chainID_list2[j].c_str(), xlen, ylen, t0, u0, TM1, TM2, TM3, TM4, TM5, rmsd0, d0_out, seqM_mat[i][j].c_str(), seqxA_mat[i][j].c_str(), seqyA_mat[i][j].c_str(), Liden, n_ali8, L_ali, TM_ali, rmsd_ali, TM_0, d0_0, d0A, d0B, Lnorm_ass, d0_scale, d0a, d0u, "", outfmt_opt, ter_opt, false, split_opt, 0, "", false, a_opt, false, d_opt, 0, resi_vec1, resi_vec2); /* clean up */ seqxA.clear(); seqM.clear(); seqyA.clear(); sequence[0].clear(); sequence[1].clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); delete[]invmap; do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); DeleteArray(&xt,xlen); } sequence.clear(); return; } void MMalign_se_final( const string xname, const string yname, const vector chainID_list1, const vector chainID_list2, string fname_super, string fname_lign, string fname_matrix, const vector > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqM_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, const double d0_scale, const bool m_opt, const int o_opt, const int outfmt_opt, const int ter_opt, const int split_opt, const bool a_opt, const bool d_opt, const bool fast_opt, const bool full_opt, const int mirror_opt, const vector&resi_vec1, const vector&resi_vec2) { int i,j; int xlen=0; int ylen=0; for (i=0;ido_vec; double Lnorm_ass=len_aa+len_na; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; int *invmap = new int[ylen+1]; /* entry function for structure alignment */ se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 3, a_opt, false, d_opt, fast_opt, mol_type, invmap); delete [] invmap; /* prepare full complex alignment */ string chainID1=""; string chainID2=""; sequence.clear(); sequence.push_back(""); // seqxA sequence.push_back(""); // seqyA sequence.push_back(""); // seqM int aln_start=0; int aln_end=0; for (i=0;i=0) continue; chainID1+=chainID_list1[i]; chainID2+=':'; string s(seqx_vec[i].begin(),seqx_vec[i].end()); sequence[0]+=s.substr(0,xlen_vec[i])+'*'; sequence[1]+=string(xlen_vec[i],'-')+'*'; s.clear(); sequence[2]+=string(xlen_vec[i],' ')+'*'; } for (j=0;j=0) continue; chainID1+=':'; chainID2+=chainID_list2[j]; string s(seqy_vec[j].begin(),seqy_vec[j].end()); sequence[0]+=string(ylen_vec[j],'-')+'*'; sequence[1]+=s.substr(0,ylen_vec[j])+'*'; s.clear(); sequence[2]+=string(ylen_vec[j],' ')+'*'; } /* print alignment */ output_results(xname, yname, chainID1.c_str(), chainID2.c_str(), xlen, ylen, t0, u0, TM1, TM2, TM3, TM4, TM5, rmsd0, d0_out, sequence[2].c_str(), sequence[0].c_str(), sequence[1].c_str(), Liden, n_ali8, L_ali, TM_ali, rmsd_ali, TM_0, d0_0, d0A, d0B, 0, d0_scale, d0a, d0u, (m_opt?fname_matrix:"").c_str(), outfmt_opt, ter_opt, true, split_opt, o_opt, fname_super, false, a_opt, false, d_opt, mirror_opt, resi_vec1, resi_vec2); /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); delete [] seqx; delete [] seqy; delete [] secx; delete [] secy; DeleteArray(&xa,xlen); DeleteArray(&ya,ylen); sequence[0].clear(); sequence[1].clear(); sequence[2].clear(); do_vec.clear(); if (!full_opt) return; if (outfmt_opt<=2) cout<<"# End of alignment for full complex. The following blocks list alignments for individual chains."<0) Lnorm_ass=len_na; sequence[0]=seqxA_mat[i][j]; sequence[1]=seqyA_mat[i][j]; /* entry function for structure alignment */ se_main(xt, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 1, a_opt, 2, d_opt, mol_vec1[i]+mol_vec2[j], 1, invmap); //TM2=TM4*Lnorm_ass/xlen; //TM1=TM4*Lnorm_ass/ylen; //d0A=d0u; //d0B=d0u; TMave_mat[i][j]=TM4*Lnorm_ass; /* print result */ if (j==assign1_list[i]) output_results(xname, yname, chainID_list1[i].c_str(), chainID_list2[j].c_str(), xlen, ylen, t0, u0, TM1, TM2, TM3, TM4, TM5, rmsd0, d0_out, seqM_mat[i][j].c_str(), seqxA_mat[i][j].c_str(), seqyA_mat[i][j].c_str(), Liden, n_ali8, L_ali, TM_ali, rmsd_ali, TM_0, d0_0, d0A, d0B, Lnorm_ass, d0_scale, d0a, d0u, "", outfmt_opt, ter_opt, false, split_opt, 0, "", false, a_opt, false, d_opt, 0, resi_vec1, resi_vec2); /* clean up */ seqxA.clear(); seqM.clear(); seqyA.clear(); sequence[0].clear(); sequence[1].clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); delete[]invmap; do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); DeleteArray(&xt,xlen); } sequence.clear(); return; } void copy_chain_assign_data(int chain1_num, int chain2_num, vector &sequence, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, double **TMave_mat, vector >&seqxA_tmp, vector >&seqyA_tmp, int *assign1_tmp, int *assign2_tmp, double **TMave_tmp) { int i,j; for (i=0;i > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, double d0_scale, bool fast_opt, map &chainmap, const int byresi_opt=0) { /* tmp assignment */ double total_score; int *assign1_tmp, *assign2_tmp; assign1_tmp=new int[chain1_num]; assign2_tmp=new int[chain2_num]; double **TMave_tmp; NewArray(&TMave_tmp,chain1_num,chain2_num); vector tmp_str_vec(chain2_num,""); vector >seqxA_tmp(chain1_num,tmp_str_vec); vector >seqyA_tmp(chain1_num,tmp_str_vec); vector sequence_tmp; copy_chain_assign_data(chain1_num, chain2_num, sequence_tmp, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat, seqxA_tmp, seqyA_tmp, assign1_tmp, assign2_tmp, TMave_tmp); for (int iter=0;iter().swap(tmp_str_vec); vector >().swap(seqxA_tmp); vector >().swap(seqyA_tmp); } /* Input: vectors x, y, rotation matrix t, u, scale factor d02, and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_TM_dimer(bool **path, double **val, double **x, double **y, int len1, int len2, bool **mask, double t[3], double u[3][3], double d02, double gap_open, int j2i[]) { int i, j; double h, v, d; //initialization for(i=0; i<=len1; i++) { //val[i][0]=0; val[i][0]=i*gap_open; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { //val[0][j]=0; val[0][j]=j*gap_open; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } double xx[3], dij; //decide matrix and path for(i=1; i<=len1; i++) { transform(t, u, &x[i-1][0], xx); for(j=1; j<=len2; j++) { d=FLT_MIN; if (mask[i][j]) { dij=dist(xx, &y[j-1][0]); d=val[i-1][j-1] + 1.0/(1+dij/d02); } //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } /* +ss * Input: secondary structure secx, secy, and gap_open * Output: j2i[1:len2] \in {1:len1} U {-1} * path[0:len1, 0:len2]=1,2,3, from diagonal, horizontal, vertical */ void NWDP_TM_dimer(bool **path, double **val, const char *secx, const char *secy, const int len1, const int len2, bool **mask, const double gap_open, int j2i[]) { int i, j; double h, v, d; //initialization for(i=0; i<=len1; i++) { //val[i][0]=0; val[i][0]=i*gap_open; path[i][0]=false; //not from diagonal } for(j=0; j<=len2; j++) { //val[0][j]=0; val[0][j]=j*gap_open; path[0][j]=false; //not from diagonal j2i[j]=-1; //all are not aligned, only use j2i[1:len2] } //decide matrix and path for(i=1; i<=len1; i++) { for(j=1; j<=len2; j++) { d=FLT_MIN; if (mask[i][j]) d=val[i-1][j-1] + 1.0*(secx[i-1]==secy[j-1]); //symbol insertion in horizontal (= a gap in vertical) h=val[i-1][j]; if(path[i-1][j]) h += gap_open; //aligned in last position //symbol insertion in vertical v=val[i][j-1]; if(path[i][j-1]) v += gap_open; //aligned in last position if(d>=h && d>=v) { path[i][j]=true; //from diagonal val[i][j]=d; } else { path[i][j]=false; //from horizontal if(v>=h) val[i][j]=v; else val[i][j]=h; } } //for i } //for j //trace back to extract the alignment i=len1; j=len2; while(i>0 && j>0) { if(path[i][j]) //from diagonal { j2i[j-1]=i-1; i--; j--; } else { h=val[i-1][j]; if(path[i-1][j]) h +=gap_open; v=val[i][j-1]; if(path[i][j-1]) v +=gap_open; if(v>=h) j--; else i--; } } } //heuristic run of dynamic programing iteratively to find the best alignment //input: initial rotation matrix t, u // vectors x and y, d0 //output: best alignment that maximizes the TMscore, will be stored in invmap double DP_iter_dimer(double **r1, double **r2, double **xtm, double **ytm, double **xt, bool **path, double **val, double **x, double **y, int xlen, int ylen, bool **mask, double t[3], double u[3][3], int invmap0[], int g1, int g2, int iteration_max, double local_d0_search, double D0_MIN, double Lnorm, double d0, double score_d8) { double gap_open[2]={-0.6, 0}; double rmsd; int *invmap=new int[ylen+1]; int iteration, i, j, k; double tmscore, tmscore_max, tmscore_old=0; int score_sum_method=8, simplify_step=40; tmscore_max=-1; //double d01=d0+1.5; double d02=d0*d0; for(int g=g1; g=0) //aligned { xtm[k][0]=x[i][0]; xtm[k][1]=x[i][1]; xtm[k][2]=x[i][2]; ytm[k][0]=y[j][0]; ytm[k][1]=y[j][1]; ytm[k][2]=y[j][2]; k++; } } tmscore = TMscore8_search(r1, r2, xtm, ytm, xt, k, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); if(tmscore>tmscore_max) { tmscore_max=tmscore; for(i=0; i0) { if(fabs(tmscore_old-tmscore)<0.000001) break; } tmscore_old=tmscore; }// for iteration }//for gapopen delete []invmap; return tmscore_max; } void get_initial_ss_dimer(bool **path, double **val, const char *secx, const char *secy, int xlen, int ylen, bool **mask, int *y2x) { double gap_open=-1.0; NWDP_TM_dimer(path, val, secx, secy, xlen, ylen, mask, gap_open, y2x); } bool get_initial5_dimer( double **r1, double **r2, double **xtm, double **ytm, bool **path, double **val, double **x, double **y, int xlen, int ylen, bool **mask, int *y2x, double d0, double d0_search, const bool fast_opt, const double D0_MIN) { double GL, rmsd; double t[3]; double u[3][3]; double d01 = d0 + 1.5; if (d01 < D0_MIN) d01 = D0_MIN; double d02 = d01*d01; double GLmax = 0; int aL = getmin(xlen, ylen); int *invmap = new int[ylen + 1]; // jump on sequence1--------------> int n_jump1 = 0; if (xlen > 250) n_jump1 = 45; else if (xlen > 200) n_jump1 = 35; else if (xlen > 150) n_jump1 = 25; else n_jump1 = 15; if (n_jump1 > (xlen / 3)) n_jump1 = xlen / 3; // jump on sequence2--------------> int n_jump2 = 0; if (ylen > 250) n_jump2 = 45; else if (ylen > 200) n_jump2 = 35; else if (ylen > 150) n_jump2 = 25; else n_jump2 = 15; if (n_jump2 > (ylen / 3)) n_jump2 = ylen / 3; // fragment to superimpose--------------> int n_frag[2] = { 20, 100 }; if (n_frag[0] > (aL / 3)) n_frag[0] = aL / 3; if (n_frag[1] > (aL / 2)) n_frag[1] = aL / 2; // start superimpose search--------------> if (fast_opt) { n_jump1*=5; n_jump2*=5; } bool flag = false; for (int i_frag = 0; i_frag < 2; i_frag++) { int m1 = xlen - n_frag[i_frag] + 1; int m2 = ylen - n_frag[i_frag] + 1; for (int i = 0; iGLmax) { GLmax = GL; for (int ii = 0; ii sequence, const double Lnorm_ass, const double d0_scale, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const bool fast_opt, const int mol_type, const double TMcut=-1) { double D0_MIN; //for d0 double Lnorm; //normalization length double score_d8,d0,d0_search,dcu0;//for TMscore search double t[3], u[3][3]; //Kabsch translation vector and rotation matrix double **score; // Input score table for dynamic programming bool **path; // for dynamic programming double **val; // for dynamic programming double **xtm, **ytm; // for TMscore search engine double **xt; //for saving the superposed version of r_1 or xtm double **r1, **r2; // for Kabsch rotation /***********************/ /* allocate memory */ /***********************/ int minlen = min(xlen, ylen); NewArray(&score, xlen+1, ylen+1); NewArray(&path, xlen+1, ylen+1); NewArray(&val, xlen+1, ylen+1); NewArray(&xtm, minlen, 3); NewArray(&ytm, minlen, 3); NewArray(&xt, xlen, 3); NewArray(&r1, minlen, 3); NewArray(&r2, minlen, 3); /***********************/ /* parameter set */ /***********************/ parameter_set4search(xlen, ylen, D0_MIN, Lnorm, score_d8, d0, d0_search, dcu0); int simplify_step = 40; //for simplified search engine int score_sum_method = 8; //for scoring method, whether only sum over pairs with dis= ylen || i1 >= xlen) kk1 = L; else if (sequence[0][kk1] != '-') invmap[i2] = i1; } } //--------------- 2. Align proteins from original alignment double prevD0_MIN = D0_MIN;// stored for later use int prevLnorm = Lnorm; double prevd0 = d0; TM_ali = standard_TMscore(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, L_ali, rmsd_ali, D0_MIN, Lnorm, d0, d0_search, score_d8, t, u, mol_type); D0_MIN = prevD0_MIN; Lnorm = prevLnorm; d0 = prevd0; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, 40, 8, local_d0_search, true, Lnorm, score_d8, d0); if (TM > TMmax) { TMmax = TM; for (i = 0; iTMmax) TMmax = TM; if (TMcut>0) copy_t_u(t, u, t0, u0); //run dynamic programing iteratively to find the best alignment TM = DP_iter_dimer(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, mask, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.5*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 2; } } /************************************************************/ /* get initial alignment based on secondary structure */ /************************************************************/ get_initial_ss_dimer(path, val, secx, secy, xlen, ylen, mask, invmap); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*0.2) { TM = DP_iter_dimer(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, mask, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.52*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 3; } } /************************************************************/ /* get initial alignment based on local superposition */ /************************************************************/ //=initial5 in original TM-align if (get_initial5_dimer( r1, r2, xtm, ytm, path, val, xa, ya, xlen, ylen, mask, invmap, d0, d0_search, fast_opt, D0_MIN)) { TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter_dimer(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, mask, t, u, invmap, 0, 2, 2, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (int i = 0; i0) copy_t_u(t, u, t0, u0); } } } else cerr << "\n\nWarning: initial alignment from local superposition fail!\n\n" << endl; if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.54*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 4; } } /********************************************************************/ /* get initial alignment by local superposition+secondary structure */ /********************************************************************/ //=initial3 in original TM-align get_initial_ssplus_dimer(r1, r2, score, path, val, secx, secy, xa, ya, xlen, ylen, mask, invmap0, invmap, D0_MIN, d0); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter_dimer(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, mask, t, u, invmap, 0, 2, (fast_opt)?2:30, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.56*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 5; } } /*******************************************************************/ /* get initial alignment based on fragment gapless threading */ /*******************************************************************/ //=initial4 in original TM-align get_initial_fgt(r1, r2, xtm, ytm, xa, ya, xlen, ylen, invmap, d0, d0_search, dcu0, fast_opt, t, u); TM = detailed_search(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, simplify_step, score_sum_method, local_d0_search, Lnorm, score_d8, d0); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } if (TM > TMmax*ddcc) { TM = DP_iter_dimer(r1, r2, xtm, ytm, xt, path, val, xa, ya, xlen, ylen, mask, t, u, invmap, 1, 2, 2, local_d0_search, D0_MIN, Lnorm, d0, score_d8); if (TM>TMmax) { TMmax = TM; for (i = 0; i0) copy_t_u(t, u, t0, u0); } } if (TMcut>0) // pre-terminate if TM-score is too low { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.58*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 6; } } //************************************************// // get initial alignment from user's input: // //************************************************// if (i_opt==1)// if input has set parameter for "-i" { for (int j = 0; j < ylen; j++)// Set aligned position to be "-1" invmap[j] = -1; int i1 = -1;// in C version, index starts from zero, not from one int i2 = -1; int L1 = sequence[0].size(); int L2 = sequence[1].size(); int L = min(L1, L2);// Get positions for aligned residues for (int kk1 = 0; kk1 < L; kk1++) { if (sequence[0][kk1] != '-') i1++; if (sequence[1][kk1] != '-') { i2++; if (i2 >= ylen || i1 >= xlen) kk1 = L; else if (sequence[0][kk1] != '-') invmap[i2] = i1; } } //--------------- 2. Align proteins from original alignment double prevD0_MIN = D0_MIN;// stored for later use int prevLnorm = Lnorm; double prevd0 = d0; TM_ali = standard_TMscore(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, L_ali, rmsd_ali, D0_MIN, Lnorm, d0, d0_search, score_d8, t, u, mol_type); D0_MIN = prevD0_MIN; Lnorm = prevLnorm; d0 = prevd0; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap, t, u, 40, 8, local_d0_search, true, Lnorm, score_d8, d0); if (TM > TMmax) { TMmax = TM; for (i = 0; iTMmax) { TMmax = TM; for (i = 0; i=0) { flag=true; break; } } if(!flag) { cout << "There is no alignment between the two structures! " << "Program stop with no result!" << endl; TM1=TM2=TM3=TM4=TM5=0; return 1; } /* last TM-score pre-termination */ if (TMcut>0) { double TMtmp=approx_TM(xlen, ylen, a_opt, xa, ya, t0, u0, invmap0, mol_type); if (TMtmp<0.6*TMcut) { TM1=TM2=TM3=TM4=TM5=TMtmp; clean_up_after_approx_TM(invmap0, invmap, score, path, val, xtm, ytm, xt, r1, r2, xlen, minlen); return 7; } } //********************************************************************// // Detailed TMscore search engine --> prepare for final TMscore // //********************************************************************// //run detailed TMscore search engine for the best alignment, and //extract the best rotation matrix (t, u) for the best alignment simplify_step=1; if (fast_opt) simplify_step=40; score_sum_method=8; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap0, t, u, simplify_step, score_sum_method, local_d0_search, false, Lnorm, score_d8, d0); //select pairs with dis=0)//aligned { n_ali++; d=sqrt(dist(&xt[i][0], &ya[j][0])); if (d <= score_d8 || (i_opt == 3)) { m1[k]=i; m2[k]=j; xtm[k][0]=xa[i][0]; xtm[k][1]=xa[i][1]; xtm[k][2]=xa[i][2]; ytm[k][0]=ya[j][0]; ytm[k][1]=ya[j][1]; ytm[k][2]=ya[j][2]; r1[k][0] = xt[i][0]; r1[k][1] = xt[i][1]; r1[k][2] = xt[i][2]; r2[k][0] = ya[j][0]; r2[k][1] = ya[j][1]; r2[k][2] = ya[j][2]; k++; } } } n_ali8=k; Kabsch(r1, r2, n_ali8, 0, &rmsd0, t, u);// rmsd0 is used for final output, only recalculate rmsd0, not t & u rmsd0 = sqrt(rmsd0 / n_ali8); //****************************************// // Final TMscore // // Please set parameters for output // //****************************************// double rmsd; simplify_step=1; score_sum_method=0; double Lnorm_0=ylen; //normalized by length of structure A parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0A=d0; d0_0=d0A; local_d0_search = d0_search; TM1 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0 = TM1; //normalized by length of structure B parameter_set4final(xlen+0.0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0B=d0; local_d0_search = d0_search; TM2 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); double Lnorm_d0; if (a_opt>0) { //normalized by average length of structures A, B Lnorm_0=(xlen+ylen)*0.5; parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0a=d0; d0_0=d0a; local_d0_search = d0_search; TM3 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM3; } if (u_opt) { //normalized by user assigned length parameter_set4final(Lnorm_ass, D0_MIN, Lnorm, d0, d0_search, mol_type); d0u=d0; d0_0=d0u; Lnorm_0=Lnorm_ass; local_d0_search = d0_search; TM4 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM4; } if (d_opt) { //scaled by user assigned d0 parameter_set4scale(ylen, d0_scale, Lnorm, d0, d0_search); d0_out=d0_scale; d0_0=d0_scale; //Lnorm_0=ylen; Lnorm_d0=Lnorm_0; local_d0_search = d0_search; TM5 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM5; } /* derive alignment from superposition */ int ali_len=xlen+ylen; //maximum length of alignment seqxA.assign(ali_len,'-'); seqM.assign( ali_len,' '); seqyA.assign(ali_len,'-'); //do_rotation(xa, xt, xlen, t, u); do_rotation(xa, xt, xlen, t0, u0); int kk=0, i_old=0, j_old=0; d=0; for(int k=0; k > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, double d0_scale, bool fast_opt) { int i,j; int xlen=0; int ylen=0; vector xlen_dimer; vector ylen_dimer; for (i=0;i().swap(xlen_dimer); vector().swap(ylen_dimer); seqx = new char[xlen+1]; secx = new char[xlen+1]; NewArray(&xa, xlen, 3); seqy = new char[ylen+1]; secy = new char[ylen+1]; NewArray(&ya, ylen, 3); int mol_type=copy_chain_pair_data(xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, chain1_num, chain2_num, seqxA_mat, seqyA_mat, assign1_list, assign2_list, sequence); /* declare variable specific to this pair of TMalign */ double t0[3], u0[3][3]; double TM1, TM2; double TM3, TM4, TM5; // for a_opt, u_opt, d_opt double d0_0, TM_0; double d0A, d0B, d0u, d0a; double d0_out=5.0; string seqM, seqxA, seqyA;// for output alignment double rmsd0 = 0.0; int L_ali; // Aligned length in standard_TMscore double Liden=0; double TM_ali, rmsd_ali; // TMscore and rmsd in standard_TMscore int n_ali=0; int n_ali8=0; double Lnorm_ass=len_aa+len_na; TMalign_dimer_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, mask, sequence, Lnorm_ass, d0_scale, 1, false, true, false, fast_opt, mol_type, -1); /* clean up TM-align */ delete [] seqx; delete [] seqy; delete [] secx; delete [] secy; DeleteArray(&xa,xlen); DeleteArray(&ya,ylen); DeleteArray(&mask,xlen+1); /* re-compute chain level alignment */ total_score=0; for (i=0;i do_vec; double Lnorm_ass=len_aa; if (mol_vec1[i]+mol_vec2[j]>0) Lnorm_ass=len_na; /* entry function for structure alignment */ se_main(xt, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 0, false, 2, false, mol_vec1[i]+mol_vec2[j], 1, invmap); /* print result */ seqxA_mat[i][j]=seqxA; seqyA_mat[i][j]=seqyA; TMave_mat[i][j]=TM4*Lnorm_ass; if (assign1_list[i]==j) { if (TM4<=0) assign1_list[i]=assign2_list[j]=-1; else total_score+=TMave_mat[i][j]; } /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); delete[]invmap; do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); DeleteArray(&xt,xlen); } return; } void MMalign_cross(double & max_total_score, const int max_iter, const vector > >&xa_vec, const vector > >&ya_vec, const vector >&seqx_vec, const vector >&seqy_vec, const vector >&secx_vec, const vector >&secy_vec, const vector &mol_vec1, const vector &mol_vec2, const vector &xlen_vec, const vector &ylen_vec, double **xa, double **ya, char *seqx, char *seqy, char *secx, char *secy, int len_aa, int len_na, int chain1_num, int chain2_num, double **TMave_mat, vector >&seqxA_mat, vector >&seqyA_mat, int *assign1_list, int *assign2_list, vector&sequence, double d0_scale, bool fast_opt, map &chainmap) { /* tmp assignment */ int *assign1_tmp, *assign2_tmp; assign1_tmp=new int[chain1_num]; assign2_tmp=new int[chain2_num]; double **TMave_tmp; NewArray(&TMave_tmp,chain1_num,chain2_num); vector tmp_str_vec(chain2_num,""); vector >seqxA_tmp(chain1_num,tmp_str_vec); vector >seqyA_tmp(chain1_num,tmp_str_vec); vector sequence_tmp; copy_chain_assign_data(chain1_num, chain2_num, sequence_tmp, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat, seqxA_tmp, seqyA_tmp, assign1_tmp, assign2_tmp, TMave_tmp); double total_score=MMalign_search(xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_tmp, seqxA_tmp, seqyA_tmp, assign1_tmp, assign2_tmp, sequence_tmp, d0_scale, fast_opt, 1); if (total_score>max_total_score) { copy_chain_assign_data(chain1_num, chain2_num, sequence, seqxA_tmp, seqyA_tmp, assign1_tmp, assign2_tmp, TMave_tmp, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat); max_total_score=total_score; } if (max_iter) MMalign_iter( max_total_score, max_iter, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_mat, seqxA_mat, seqyA_mat, assign1_list, assign2_list, sequence, d0_scale, fast_opt, chainmap); /* clean up everything */ delete [] assign1_tmp; delete [] assign2_tmp; DeleteArray(&TMave_tmp,chain1_num); vector().swap(tmp_str_vec); vector >().swap(seqxA_tmp); vector >().swap(seqyA_tmp); vector().swap(sequence_tmp); return; } /* return the number of chains that are trimmed */ int trimComplex(vector > >&a_trim_vec, vector >&seq_trim_vec, vector >&sec_trim_vec, vector&len_trim_vec, const vector > >&a_vec, const vector >&seq_vec, const vector >&sec_vec, const vector &len_vec, const vector &mol_vec, const int Lchain_aa_max, const int Lchain_na_max) { int trim_chain_count=0; int chain_num=a_vec.size(); int i,j; int r1,r2; double dinter; double dinter_min; vector >dinter_vec; vector include_vec; vector seq_empty; vector > a_empty; vector xcoor(3,0); vector ycoor(3,0); int xlen,ylen; int Lchain_max; double expand=2; for (i=0;i0) Lchain_max=Lchain_na_max*expand; else Lchain_max=Lchain_aa_max*expand; if (Lchain_max<3) Lchain_max=3; if (xlen<=Lchain_max || xlen<=3) { a_trim_vec.push_back(a_vec[i]); seq_trim_vec.push_back(seq_vec[i]); sec_trim_vec.push_back(sec_vec[i]); len_trim_vec.push_back(xlen); continue; } trim_chain_count++; for (r1=0;r1 >().swap(dinter_vec); vector().swap(include_vec); vector ().swap(xcoor); vector ().swap(ycoor); return trim_chain_count; } void output_dock_rotation_matrix(const char* fname_matrix, const vector&xname_vec, const vector&yname_vec, double ** ut_mat, int *assign1_list) { fstream fout; fout.open(fname_matrix, ios::out | ios::trunc); if (fout)// succeed { int i,k; for (i=0;i=0) // previous SSE end { endi=i; for (j=starti;j') starti=i; else starti=-1; } prev_ss=secx[i]; } if (starti>=0) // previous SSE end { endi=i; for (j=starti;j > close_idx_vec(xlen, make_pair(0,0)); int i,j,k; for (i=0;i >().swap(close_idx_vec); DeleteArray(&score, xlen+1); } /* check if pairing i to j conform to sequantiality within the SSE */ inline bool sec2sq(const int i, const int j, int **secx_bond, int **secy_bond, int *fwdmap, int *invmap) { if (i<0 || j<0) return true; int ii,jj; if (secx_bond[i][0]>=0) { for (ii=secx_bond[i][0];ii=0 && (i-ii)*(j-jj)<=0) return false; } } if (secy_bond[j][0]>=0) { for (jj=secy_bond[j][0];jj=0 && (i-ii)*(j-jj)<=0) return false; } } return true; } void soi_egs(double **score, const int xlen, const int ylen, int *invmap, int **secx_bond, int **secy_bond, const int mm_opt) { int i,j; int *fwdmap=new int[xlen]; // j=fwdmap[i]; for (i=0; i=0) fwdmap[i]=j; } /* stage 1 - make initial assignment, starting from the highest score pair */ double max_score; int maxi,maxj; while(1) { max_score=0; maxi=maxj=-1; for (i=0;i=0) continue; for (j=0;j=0 || score[i+1][j+1]<=max_score) continue; if (mm_opt==6 && !sec2sq(i,j,secx_bond,secy_bond, fwdmap,invmap)) continue; maxi=i; maxj=j; max_score=score[i+1][j+1]; } } if (maxi<0) break; // no assignment; invmap[maxj]=maxi; fwdmap[maxi]=maxj; } double total_score=0; for (j=0;j=0) total_score+=score[i+1][j+1]; } /* stage 2 - swap assignment until total score cannot be improved */ int iter; int oldi,oldj; double delta_score; for (iter=0; iter prepare for final TMscore // //********************************************************************// //run detailed TMscore search engine for the best alignment, and //extract the best rotation matrix (t, u) for the best alignment simplify_step=1; if (fast_opt) simplify_step=40; score_sum_method=8; TM = detailed_search_standard(r1, r2, xtm, ytm, xt, xa, ya, xlen, ylen, invmap0, t, u, simplify_step, score_sum_method, local_d0_search, false, Lnorm, score_d8, d0); double rmsd; simplify_step=1; score_sum_method=0; double Lnorm_0=ylen; //select pairs with dis=0)//aligned { n_ali++; d=sqrt(dist(&xt[i][0], &ya[j][0])); if (d <= score_d8) { m1[k]=i; m2[k]=j; xtm[k][0]=xa[i][0]; xtm[k][1]=xa[i][1]; xtm[k][2]=xa[i][2]; ytm[k][0]=ya[j][0]; ytm[k][1]=ya[j][1]; ytm[k][2]=ya[j][2]; r1[k][0] = xt[i][0]; r1[k][1] = xt[i][1]; r1[k][2] = xt[i][2]; r2[k][0] = ya[j][0]; r2[k][1] = ya[j][1]; r2[k][2] = ya[j][2]; k++; } else fwdmap0[i]=-1; } } n_ali8=k; Kabsch(r1, r2, n_ali8, 0, &rmsd0, t, u);// rmsd0 is used for final output, only recalculate rmsd0, not t & u rmsd0 = sqrt(rmsd0 / n_ali8); //normalized by length of structure A parameter_set4final(xlen+0.0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0B=d0; local_d0_search = d0_search; TM2 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t, u, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); //****************************************// // Final TMscore 2 // //****************************************// do_rotation(xa, xt, xlen, t0, u0); k=0; for (j=0; j=0)//aligned { d=sqrt(dist(&xt[i][0], &ya[j][0])); if (d <= score_d8) { m1[k]=i; m2[k]=j; xtm[k][0]=xa[i][0]; xtm[k][1]=xa[i][1]; xtm[k][2]=xa[i][2]; ytm[k][0]=ya[j][0]; ytm[k][1]=ya[j][1]; ytm[k][2]=ya[j][2]; r1[k][0] = xt[i][0]; r1[k][1] = xt[i][1]; r1[k][2] = xt[i][2]; r2[k][0] = ya[j][0]; r2[k][1] = ya[j][1]; r2[k][2] = ya[j][2]; k++; } else invmap[j]=invmap0[j]=-1; } } //normalized by length of structure B parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0A=d0; d0_0=d0A; local_d0_search = d0_search; TM1 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0 = TM1; if (a_opt>0) { //normalized by average length of structures A, B Lnorm_0=(xlen+ylen)*0.5; parameter_set4final(Lnorm_0, D0_MIN, Lnorm, d0, d0_search, mol_type); d0a=d0; d0_0=d0a; local_d0_search = d0_search; TM3 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM3; } if (u_opt) { //normalized by user assigned length parameter_set4final(Lnorm_ass, D0_MIN, Lnorm, d0, d0_search, mol_type); d0u=d0; d0_0=d0u; Lnorm_0=Lnorm_ass; local_d0_search = d0_search; TM4 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM4; } if (d_opt) { //scaled by user assigned d0 parameter_set4scale(ylen, d0_scale, Lnorm, d0, d0_search); d0_out=d0_scale; d0_0=d0_scale; //Lnorm_0=ylen; local_d0_search = d0_search; TM5 = TMscore8_search(r1, r2, xtm, ytm, xt, n_ali8, t0, u0, simplify_step, score_sum_method, &rmsd, local_d0_search, Lnorm, score_d8, d0); TM_0=TM5; } /* derive alignment from superposition */ int ali_len=xlen+ylen; for (j=0;j=0); seqxA.assign(ali_len,'-'); seqM.assign( ali_len,' '); seqyA.assign(ali_len,'-'); //do_rotation(xa, xt, xlen, t, u); do_rotation(xa, xt, xlen, t0, u0); Liden=0; //double SO=0; for (j=0;j=0) continue; seqxA[ylen+k]=seqx[i]; k++; } //cout< &tu_tmp) { int i,j,k; for (i=0;i<3;i++) tu_tmp[i]=t0[i]; k=3; for (i=0;i<3;i++) for (j=0;j<3;j++) { tu_tmp[k]=u0[i][j]; k++; } } void tu2t_u(vector tu_tmp, double t0[3],double u0[3][3]) { int i,j,k; for (i=0;i<3;i++) t0[i]=tu_tmp[i]; k=3; for (i=0;i<3;i++) for (j=0;j<3;j++) { u0[i][j]=tu_tmp[k]; k++; } } int flexalign_main(double **xa, double **ya, const char *seqx, const char *seqy, const char *secx, const char *secy, double t0[3], double u0[3][3], vector >&tu_vec, double &TM1, double &TM2, double &TM3, double &TM4, double &TM5, double &d0_0, double &TM_0, double &d0A, double &d0B, double &d0u, double &d0a, double &d0_out, string &seqM, string &seqxA, string &seqyA, vector&do_vec, double &rmsd0, int &L_ali, double &Liden, double &TM_ali, double &rmsd_ali, int &n_ali, int &n_ali8, const int xlen, const int ylen, const vector sequence, const double Lnorm_ass, const double d0_scale, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const bool fast_opt, const int mol_type, const int hinge_opt) { vector tu_tmp(12,0); int round2=tu_vec.size(); if (round2==0) { TMalign_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, fast_opt, mol_type); t_u2tu(t0,u0,tu_tmp); tu_vec.push_back(tu_tmp); } int i,j,r; int* invmap=new int[ylen+1]; for (j=0;jTM2_h)?TM1_h:TM2_h; double TM =(TM1 >TM2 )?TM1 :TM2 ; if (TM_h>TM) { TM1=TM1_h; TM2=TM2_h; TM3=TM3_h; TM4=TM4_h; TM5=TM5_h; seqM=seqM_h; seqxA=seqxA_h; seqyA=seqyA_h; rmsd0=rmsd0_h; n_ali=n_ali_h; n_ali8=n_ali8_h; for (j=0;j r1toi(xlen_h,0); vector r2toj(ylen_h,0); int r1,r2; i=j=-1; r1=r2=0; for (r=0;r=5) { TM1=TM1_h; TM2=TM2_h; TM3=TM3_h; TM4=TM4_h; TM5=TM5_h; seqM=seqM_h; seqxA=seqxA_h; seqyA=seqyA_h; rmsd0=rmsd0_h; n_ali=n_ali_h; n_ali8=n_ali8_h; t_u2tu(t0,u0,tu_tmp); tu_vec.push_back(tu_tmp); for (j=0;jhinge="<=0) cout<<"("< seqM_char(ylen,' '); vector di_vec(ylen,-1); double d; for (hinge=tu_vec.size()-1;hinge>=0;hinge--) { tu2t_u(tu_vec[hinge],t0,u0); do_rotation(xa, xt, xlen, t0, u0); for (j=0;j=0;hinge--) { j=-1; for (r=0;r0 && (seqM[r-1]==hinge+'0' || seqM[r-1]==' ')) continue; if (r0 && seqM[r-1]!=seqM[r+1]) continue; if (r>0) seqM[r]=seqM_char[j]=seqM[r-1]; else seqM[r]=seqM_char[j]=seqM[r+1]; } } /* smooth out AFP assignment: remove singleton at the end of fragment */ char left_hinge=' '; char right_hinge=' '; for (hinge=tu_vec.size()-1;hinge>=0;hinge--) { j=-1; for (r=0;r0 && seqM[r-1]==' ' && r=0;i--) { if (seqM[i]==' ') continue; left_hinge=seqM[i]; break; } if (left_hinge==hinge+'0') continue; right_hinge=' '; for (i=r+1;i=0;hinge--) { j=-1; for (r=0;r0 && (seqM[r-1]==' ' || seqM[r-1]==hinge+'0')) continue; if (r0 && seqM[r-1]!=seqM[r+2]) continue; if (r>0) seqM[r]=seqM_char[j]=seqM[r+1]=seqM_char[j+1]=seqM[r-1]; else seqM[r]=seqM_char[j]=seqM[r+1]=seqM_char[j+1]=seqM[r+2]; } } /* smooth out AFP assignment: remove disconnected singleton */ int i1,i2; for (hinge=tu_vec.size()-1;hinge>=0;hinge--) { j=-1; for (r=0;r=0;i--) { if (seqM[i]==' ') continue; left_hinge=seqM[i]; i1=(r-i); break; } if (left_hinge==hinge+'0') continue; right_hinge=' '; for (i=r+1;i=0;hinge--) { tu2t_u(tu_vec[hinge],t0,u0); do_rotation(xa, xt, xlen, t0, u0); for (j=0;j0;hinge--) { int afp_len=0; for (r=0;r >&tu_vec, double t[3], double u[3][3]) { stringstream ss; char dest[1000]; for (int hinge=0;hinge >&tu_vec, double t[3], double u[3][3], const int ter_opt, const int mm_opt, const int split_opt, const int mirror_opt, const char *seqM, const char *seqxA, const char *seqyA, const vector&resi_vec1, const vector&resi_vec2, const string chainID1, const string chainID2, const int xlen, const int ylen, const double d0A, const int n_ali8, const double rmsd, const double TM1, const double Liden) { stringstream buf; stringstream buf_all; stringstream buf_atm; stringstream buf_all_atm; stringstream buf_all_atm_lig; //stringstream buf_pdb; stringstream buf_tm; string line; double x[3]; // before transform double x1[3]; // after transform bool after_ter; // true if passed the "TER" line in PDB string asym_id; // chain ID map resi2hinge_dict; int r,i,j; j=-1; char hinge_char=0; int ali_len=strlen(seqM); for (r=0;r=0 && seqM[r-i]!=' ') hinge_char=seqM[r-i]; else if (r+i=1) // align one chain from model 1 { chain1_sele=chainID1.substr(1); chain2_sele=chainID2.substr(1); } else if (split_opt==2 && ter_opt==0) // align one chain from each model { for (i=1;i _atom_site; int atom_site_pos; vector line_vec; string atom; // 4-character atom name string AA; // 3-character residue name string inscode; // 1-character insertion code string model_index; // model index bool is_mmcif=false; /* used for CONECT record of chain1 */ int ca_idx1=0; // all CA atoms int lig_idx1=0; // all atoms vector idx_vec; /* used for CONECT record of chain2 */ int ca_idx2=0; // all CA atoms int lig_idx2=0; // all atoms /* extract aligned region */ vector resi_aln1; vector resi_aln2; int i1=-1; int i2=-1; if (!mm_opt) { for (i=0;i=3 && line.compare(0,3,"TER")==0) after_ter=true; if (is_mmcif==false && line.size()>=54 && (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0)) // PDB format { if (line[16]!='A' && line[16]!=' ') continue; x[0]=atof(line.substr(30,8).c_str()); x[1]=atof(line.substr(38,8).c_str()); x[2]=atof(line.substr(46,8).c_str()); if (mirror_opt) x[2]=-x[2]; if (read_resi==1) resi=line.substr(22,5); else resi=line.substr(22,5)+line[21]; hinge=0; if (resi2hinge_dict.count(resi)) hinge=resi2hinge_dict[resi]; tu2t_u(tu_vec[hinge],t,u); transform(t, u, x, x1); //buf_pdb<=2) { if (ca_idx1 && asym_id.size() && asym_id!=line.substr(21,1)) { after_ter=true; continue; } asym_id=line[21]; } buf_all_atm<<"ATOM "<=2) { if (_atom_site.count("auth_asym_id")) asym_id=line_vec[_atom_site["auth_asym_id"]]; else asym_id=line_vec[_atom_site["label_asym_id"]]; if (asym_id==".") asym_id=" "; resi+=asym_id[0]; } hinge=0; if (resi2hinge_dict.count(resi)) hinge=resi2hinge_dict[resi]; tu2t_u(tu_vec[hinge],t,u); transform(t, u, x, x1); if (_atom_site.count("label_alt_id")==0 || line_vec[_atom_site["label_alt_id"]]=="." || line_vec[_atom_site["label_alt_id"]]=="A") { atom=line_vec[_atom_site["label_atom_id"]]; if (atom[0]=='"') atom=atom.substr(1); if (atom.size() && atom[atom.size()-1]=='"') atom=atom.substr(0,atom.size()-1); if (atom.size()==0) atom=" "; else if (atom.size()==1) atom=" "+atom+" "; else if (atom.size()==2) atom=" "+atom+" "; else if (atom.size()==3) atom=" "+atom; else if (atom.size()>=5) atom=atom.substr(0,4); AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) AA=AA.substr(0,3); if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; while (resi.size()<4) resi=' '+resi; if (resi.size()>4) resi=resi.substr(0,4); inscode=' '; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") inscode=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; if (_atom_site.count("auth_asym_id")) { if (chain1_sele.size()) after_ter =line_vec[_atom_site["auth_asym_id"]]!=chain1_sele; else if (ter_opt>=2 && ca_idx1 && asym_id.size() && asym_id!=line_vec[_atom_site["auth_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["auth_asym_id"]]; } else if (_atom_site.count("label_asym_id")) { if (chain1_sele.size()) after_ter =line_vec[_atom_site["label_asym_id"]]!=chain1_sele; if (ter_opt>=2 && ca_idx1 && asym_id.size() && asym_id!=line_vec[_atom_site["label_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["label_asym_id"]]; } //buf_pdb<=1 && line.compare(0,3,"END")==0) break; } } fin.close(); if (!mm_opt) buf<<"TER\n"; buf_all<<"TER\n"; if (!mm_opt) buf_atm<<"TER\n"; buf_all_atm<<"TER\n"; buf_all_atm_lig<<"TER\n"; for (i=1;i=3 && line.compare(0,3,"TER")==0) after_ter=true; if (line.size()>=54 && (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0)) // PDB format { if (line[16]!='A' && line[16]!=' ') continue; if (after_ter && line.compare(0,6,"ATOM ")==0) continue; lig_idx2++; buf_all_atm_lig<=2) { if (ca_idx2 && asym_id.size() && asym_id!=line.substr(21,1)) { after_ter=true; continue; } asym_id=line[21]; } buf_all_atm<<"ATOM "<=5) atom=atom.substr(0,4); AA=line_vec[_atom_site["label_comp_id"]]; // residue name if (AA.size()==1) AA=" "+AA; else if (AA.size()==2) AA=" " +AA; else if (AA.size()>=4) AA=AA.substr(0,3); if (_atom_site.count("auth_seq_id")) resi=line_vec[_atom_site["auth_seq_id"]]; else resi=line_vec[_atom_site["label_seq_id"]]; while (resi.size()<4) resi=' '+resi; if (resi.size()>4) resi=resi.substr(0,4); inscode=' '; if (_atom_site.count("pdbx_PDB_ins_code") && line_vec[_atom_site["pdbx_PDB_ins_code"]]!="?") inscode=line_vec[_atom_site["pdbx_PDB_ins_code"]][0]; if (_atom_site.count("auth_asym_id")) { if (chain2_sele.size()) after_ter =line_vec[_atom_site["auth_asym_id"]]!=chain2_sele; if (ter_opt>=2 && ca_idx2 && asym_id.size() && asym_id!=line_vec[_atom_site["auth_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["auth_asym_id"]]; } else if (_atom_site.count("label_asym_id")) { if (chain2_sele.size()) after_ter =line_vec[_atom_site["label_asym_id"]]!=chain2_sele; if (ter_opt>=2 && ca_idx2 && asym_id.size() && asym_id!=line_vec[_atom_site["label_asym_id"]]) after_ter=true; asym_id=line_vec[_atom_site["label_asym_id"]]; } if (after_ter==false || line_vec[_atom_site["group_PDB"]]=="HETATM") { lig_idx2++; buf_all_atm_lig<=1 && line.compare(0,3,"END")==0) break; } } fin.close(); if (!mm_opt) buf<<"TER\n"; buf_all<<"TER\n"; if (!mm_opt) buf_atm<<"TER\n"; buf_all_atm<<"TER\n"; buf_all_atm_lig<<"TER\n"; for (i=ca_idx1+1;i pml_list; pml_list.push_back(fname_super+""); pml_list.push_back(fname_super+"_atm"); pml_list.push_back(fname_super+"_all"); pml_list.push_back(fname_super+"_all_atm"); pml_list.push_back(fname_super+"_all_atm_lig"); for (i=0;i >&tu_vec, double t[3], double u[3][3], const int ter_opt, const int mm_opt, const int split_opt, const int mirror_opt, const char *seqM, const char *seqxA, const char *seqyA, const vector&resi_vec1, const vector&resi_vec2, const string chainID1, const string chainID2) { int compress_type=0; // uncompressed file ifstream fin; #ifndef REDI_PSTREAM_H_SEEN ifstream fin_gz; #else redi::ipstream fin_gz; // if file is compressed if (xname.size()>=3 && xname.substr(xname.size()-3,3)==".gz") { fin_gz.open("gunzip -c "+xname); compress_type=1; } else if (xname.size()>=4 && xname.substr(xname.size()-4,4)==".bz2") { fin_gz.open("bzcat "+xname); compress_type=2; } else #endif fin.open(xname.c_str()); map resi2hinge_dict; int r,i,j; j=-1; char hinge_char=0; int xlen=resi_vec1.size(); int ali_len=strlen(seqM); for (r=0;r=0 && seqM[r-i]!=' ') hinge_char=seqM[r-i]; else if (r+i _atom_site; size_t atom_site_pos; vector line_vec; int infmt=-1; // 0 - PDB, 3 - PDBx/mmCIF int hinge=0; string asym_id="."; // this is similar to chainID, except that // chainID is char while asym_id is a string // with possibly multiple char while (compress_type?fin_gz.good():fin.good()) { if (compress_type) getline(fin_gz, line); else getline(fin, line); if (line.compare(0, 6, "ATOM ")==0 || line.compare(0, 6, "HETATM")==0) // PDB format { infmt=0; x[0]=atof(line.substr(30,8).c_str()); x[1]=atof(line.substr(38,8).c_str()); x[2]=atof(line.substr(46,8).c_str()); if (mirror_opt) x[2]=-x[2]; if (read_resi==1) resi=line.substr(22,5); else resi=line.substr(22,5)+line[21]; hinge=0; if (resi2hinge_dict.count(resi)) hinge=resi2hinge_dict[resi]; tu2t_u(tu_vec[hinge],t,u); transform(t, u, x, x1); buf<=2) { if (_atom_site.count("auth_asym_id")) asym_id=line_vec[_atom_site["auth_asym_id"]]; else asym_id=line_vec[_atom_site["label_asym_id"]]; if (asym_id==".") asym_id=" "; resi+=asym_id[0]; } hinge=0; if (resi2hinge_dict.count(resi)) hinge=resi2hinge_dict[resi]; tu2t_u(tu_vec[hinge],t,u); transform(t, u, x, x1); for (atom_site_pos=0; atom_site_pos<_atom_site.size(); atom_site_pos++) { if (atom_site_pos==_atom_site["Cartn_x"]) buf<=1 && line.compare(0,3,"END")==0) break; } } if (compress_type) fin_gz.close(); else fin.close(); string fname_super_full=fname_super; if (infmt==0) fname_super_full+=".pdb"; else if (infmt==3) fname_super_full+=".cif"; ofstream fp; fp.open(fname_super_full.c_str()); fp<=1) // align one chain from model 1 { chain1_sele=" and c. "+chainID1.substr(1); chain2_sele=" and c. "+chainID2.substr(1); } else if (split_opt==2 && ter_opt==0) // align one chain from each model { for (i=1;i pml_list; pml_list.push_back(fname_super+""); pml_list.push_back(fname_super+"_atm"); pml_list.push_back(fname_super+"_all"); pml_list.push_back(fname_super+"_all_atm"); pml_list.push_back(fname_super+"_all_atm_lig"); for (int p=0;p >&tu_vec, const double TM1, const double TM2, const double TM3, const double TM4, const double TM5, const double rmsd, const double d0_out, const char *seqM, const char *seqxA, const char *seqyA, const double Liden, const int n_ali8, const int L_ali, const double TM_ali, const double rmsd_ali, const double TM_0, const double d0_0, const double d0A, const double d0B, const double Lnorm_ass, const double d0_scale, const double d0a, const double d0u, const char* fname_matrix, const int outfmt_opt, const int ter_opt, const int mm_opt, const int split_opt, const int o_opt, const string fname_super, const int i_opt, const int a_opt, const bool u_opt, const bool d_opt, const int mirror_opt, const vector&resi_vec1, const vector&resi_vec2) { if (outfmt_opt<=0) { printf("\nName of Structure_1: %s%s (to be superimposed onto Structure_2)\n", xname.c_str(), chainID1.c_str()); printf("Name of Structure_2: %s%s\n", yname.c_str(), chainID2.c_str()); printf("Length of Structure_1: %d residues\n", xlen); printf("Length of Structure_2: %d residues\n\n", ylen); if (i_opt) printf("User-specified initial alignment: TM/Lali/rmsd = %7.5lf, %4d, %6.3lf\n", TM_ali, L_ali, rmsd_ali); printf("Aligned length= %d, RMSD= %6.2f, Seq_ID=n_identical/n_aligned= %4.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); printf("TM-score= %6.5f (normalized by length of Structure_1: L=%d, d0=%.2f)\n", TM2, xlen, d0B); printf("TM-score= %6.5f (normalized by length of Structure_2: L=%d, d0=%.2f)\n", TM1, ylen, d0A); if (a_opt==1) printf("TM-score= %6.5f (if normalized by average length of two structures: L=%.1f, d0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if (u_opt) printf("TM-score= %6.5f (normalized by user-specified L=%.2f and d0=%.2f)\n", TM4, Lnorm_ass, d0u); if (d_opt) printf("TM-score= %6.5f (scaled by user-specified d0=%.2f, and L=%d)\n", TM5, d0_scale, ylen); printf("(You should use TM-score normalized by length of the reference structure)\n"); //output alignment printf("\n([0-9] denote different aligned fragment pairs separated by different hinges)\n"); printf("%s\n", seqxA); printf("%s\n", seqM); printf("%s\n", seqyA); } else if (outfmt_opt==1) { printf(">%s%s\tL=%d\td0=%.2f\tseqID=%.3f\tTM-score=%.5f\n", xname.c_str(), chainID1.c_str(), xlen, d0B, Liden/xlen, TM2); printf("%s\n", seqxA); printf(">%s%s\tL=%d\td0=%.2f\tseqID=%.3f\tTM-score=%.5f\n", yname.c_str(), chainID2.c_str(), ylen, d0A, Liden/ylen, TM1); printf("%s\n", seqyA); printf("# Lali=%d\tRMSD=%.2f\tseqID_ali=%.3f\n", n_ali8, rmsd, (n_ali8>0)?Liden/n_ali8:0); if (i_opt) printf("# User-specified initial alignment: TM=%.5lf\tLali=%4d\trmsd=%.3lf\n", TM_ali, L_ali, rmsd_ali); if(a_opt) printf("# TM-score=%.5f (normalized by average length of two structures: L=%.1f\td0=%.2f)\n", TM3, (xlen+ylen)*0.5, d0a); if(u_opt) printf("# TM-score=%.5f (normalized by user-specified L=%.2f\td0=%.2f)\n", TM4, Lnorm_ass, d0u); if(d_opt) printf("# TM-score=%.5f (scaled by user-specified d0=%.2f\tL=%d)\n", TM5, d0_scale, ylen); printf("$$$$\n"); } else if (outfmt_opt==2) { printf("%s%s\t%s%s\t%.4f\t%.4f\t%.2f\t%4.3f\t%4.3f\t%4.3f\t%d\t%d\t%d", xname.c_str(), chainID1.c_str(), yname.c_str(), chainID2.c_str(), TM2, TM1, rmsd, Liden/xlen, Liden/ylen, (n_ali8>0)?Liden/n_ali8:0, xlen, ylen, n_ali8); } cout << endl; if (strlen(fname_matrix)) output_flexalign_rotation_matrix( fname_matrix, tu_vec, t, u); if (o_opt==1) output_flexalign_pymol(xname, yname, fname_super, tu_vec, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2); else if (o_opt==2) output_flexalign_rasmol(xname, yname, fname_super, tu_vec, t, u, ter_opt, mm_opt, split_opt, mirror_opt, seqM, seqxA, seqyA, resi_vec1, resi_vec2, chainID1, chainID2, xlen, ylen, d0A, n_ali8, rmsd, TM1, Liden); } /* USalign.cpp */ /* command line argument parsing and document of US-align main program */ void print_version() { cout << "\n" " ********************************************************************\n" " * US-align (Version 20240730) *\n" " * Universal Structure Alignment of Proteins and Nucleic Acids *\n" " * Reference: C Zhang, M Shine, AM Pyle, Y Zhang. (2022) Nat Methods*\n" " * C Zhang, AM Pyle (2022) iScience. *\n" " * Please email comments and suggestions to zhang@zhanggroup.org *\n" " ********************************************************************" << endl; } void print_extra_help() { cout << "Additional options:\n" " -v Print the version of US-align\n" "\n" " -a TM-score normalized by the average length of two structures\n" " T or F, (default F). -a does not change the final alignment.\n" "\n" " -fast Fast but slightly inaccurate alignment\n" "\n" " -dir Perform all-against-all alignment among the list of PDB\n" " chains listed by 'chain_list' under 'chain_folder'.\n" " $ USalign -dir chain_folder/ chain_list\n" "\n" //"-dirpair Perform batch alignment for each pair of chains listed by\n" //" 'chain_pair_list' under 'chain_folder'. Each line consist of\n" //" two chains, separated by tab or space.\n" //" $ USalign -dirpair chain_folder/ chain_pair_list\n" //"\n" " -dir1 Use chain2 to search a list of PDB chains listed by 'chain1_list'\n" " under 'chain1_folder'.\n" " $ USalign -dir1 chain1_folder/ chain1_list chain2\n" "\n" " -dir2 Use chain1 to search a list of PDB chains listed by 'chain2_list'\n" " under 'chain2_folder'\n" " $ USalign chain1 -dir2 chain2_folder/ chain2_list\n" "\n" " -suffix (Only when -dir1 and/or -dir2 are set, default is empty)\n" " add file name suffix to files listed by chain1_list or chain2_list\n" "\n" " -atom 4-character atom name used to represent a residue.\n" " Default is \" C3'\" for RNA/DNA and \" CA \" for proteins\n" " (note the spaces before and after CA).\n" "\n" " -split Whether to split PDB file into multiple chains\n" " 0: treat the whole structure as one single chain\n" " 1: treat each MODEL as a separate chain\n" " 2: (default) treat each chain as a separate chain\n" "\n" " -outfmt Output format\n" " 0: (default) full output\n" " 1: fasta format compact output\n" " 2: tabular format very compact output\n" " -1: full output, but without version or citation information\n" "\n" " -TMcut -1: (default) do not consider TMcut\n" " Values in [0.5,1): Do not proceed with TM-align for this\n" " structure pair if TM-score is unlikely to reach TMcut.\n" " TMcut is normalized as set by -a option:\n" " -2: normalized by longer structure length\n" " -1: normalized by shorter structure length\n" " 0: (default, same as F) normalized by second structure\n" " 1: same as T, normalized by average structure length\n" "\n" " -mirror Whether to align the mirror image of input structure\n" " 0: (default) do not align mirrored structure\n" " 1: align mirror of Structure_1 to origin Structure_2,\n" " which usually requires the '-het 1' option:\n" " $ USalign 4glu.pdb 3p9w.pdb -mirror 1 -het 1\n" "\n" " -het Whether to align residues marked as 'HETATM' in addition to 'ATOM '\n" " 0: (default) only align 'ATOM ' residues\n" " 1: align both 'ATOM ' and 'HETATM' residues\n" " 2: align both 'ATOM ' and MSE residues\n" "\n" " -full Whether to show full pairwise alignment of individual chains for\n" " -mm 2 or 4. T or F, (default F)\n" //"\n" //" -closeK Number of closest atoms used for sequence order independent\n" //" initial alignment. default: 5\n" //"\n" //" -hinge Maximum number of hinge allowed in flexible alignment. default: 9\n" "\n" " -se Do not perform superposition. Useful for extracting alignment from\n" " superposed structure pairs\n" "\n" " -infmt1 Input format for structure_1\n" " -infmt2 Input format for structure_2\n" " -1: (default) automatically detect PDB or PDBx/mmCIF format\n" " 0: PDB format\n" " 1: SPICKER format\n" //" 2: xyz format\n" " 3: PDBx/mmCIF format\n" "\n" "-chainmap (only useful for -mm 1) use the final chain mapping 'chainmap.txt'\n" " specified by user. 'chainmap.txt' is a tab-seperated text with two\n" " columns, one for each complex\n" "\n" "-chain1 Chains to parse in structure_1\n" "-chain2 Chains to parse in structure_2. Use _ for a chain without chain ID.\n" " Multiple chains can be separated by commas, e.g.,\n" " USalign -chain1 C,D,E,F 5jdo.pdb -chain2 A,B,C,D 3wtg.pdb -ter 0\n" "\n" "Advanced usage 1 (generate an image for a pair of superposed structures):\n" " USalign 1cpc.pdb 1mba.pdb -o sup\n" " pymol -c -d @sup_all_atm.pml -g sup_all_atm.png\n" "\n" "Advanced usage 2 (a quick search of query.pdb against I-TASSER PDB library):\n" " wget https://zhanggroup.org/library/PDB.tar.bz2\n" " tar -xjvf PDB.tar.bz2\n" " USalign query.pdb -dir2 PDB/ PDB/list -suffix .pdb -outfmt 2 -fast\n" <= length\n" " of protein to avoid TM-score >1. -u does not change final alignment.\n" "\n" " -o Output superposed structure1 to sup.* for PyMOL viewing.\n" " $ USalign structure1.pdb structure2.pdb -o sup\n" " $ pymol -d @sup.pml # C-alpha trace aligned region\n" " $ pymol -d @sup_all.pml # C-alpha trace whole chain\n" " $ pymol -d @sup_atm.pml # full-atom aligned region\n" " $ pymol -d @sup_all_atm.pml # full-atom whole chain\n" " $ pymol -d @sup_all_atm_lig.pml # full-atom with all molecules\n" "\n" " -rasmol Output superposed structure1 to sup.* for RasMol viewing.\n" " $ USalign structure1.pdb structure2.pdb -rasmol sup\n" " $ rasmol -script sup # C-alpha trace aligned region\n" " $ rasmol -script sup_all # C-alpha trace whole chain\n" " $ rasmol -script sup_atm # full-atom aligned region\n" " $ rasmol -script sup_all_atm # full-atom whole chain\n" " $ rasmol -script sup_all_atm_lig # full-atom with all molecules\n" "\n" " -do Output distance of aligned residue pairs\n" "\n" //" -h Print the full help message, including additional options\n" //"\n" "Example usages ('gunzip' program is needed to read .gz compressed files):\n" " USalign 101m.cif.gz 1mba.pdb # pairwise monomeric protein alignment\n" " USalign 1qf6.cif 5yyn.pdb.gz -mol RNA # pairwise monomeric RNA alignment\n" " USalign model.pdb native.pdb -TMscore 1 # calculate TM-score between two conformations of a monomer\n" " USalign 4v4a.cif 4v49.cif -mm 1 -ter 1 # oligomeric alignment for asymmetic units\n" " USalign 3ksc.pdb1 4lej.pdb1 -mm 1 -ter 0 # oligomeric alignment for biological units\n" " USalign 1ajk.pdb.gz 2ayh.pdb.gz -mm 3 # circular permutation alignment\n" < &sequence, const double Lnorm_ass, const double d0_scale, const bool m_opt, const int i_opt, const int o_opt, const int a_opt, const bool u_opt, const bool d_opt, const double TMcut, const int infmt1_opt, const int infmt2_opt, const int ter_opt, const int split_opt, const int outfmt_opt, const bool fast_opt, const int cp_opt, const int mirror_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir_opt, const string &dirpair_opt, const string &dir1_opt, const string &dir2_opt, const vector &chain2parse1, const vector &chain2parse2, const vector &model2parse1, const vector &model2parse2, const int byresi_opt, const vector &chain1_list, const vector &chain2_list, const bool se_opt, const bool do_opt) { /* declare previously global variables */ vector >PDB_lines1; // text of chain1 vector >PDB_lines2; // text of chain2 vector mol_vec1; // molecule type of chain1, RNA if >0 vector mol_vec2; // molecule type of chain2, RNA if >0 vector chainID_list1; // list of chainID1 vector chainID_list2; // list of chainID2 int i,j; // file index int chain_i,chain_j; // chain index int r; // residue index int xlen, ylen; // chain length int xchainnum,ychainnum;// number of chains in a PDB file char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure double **xa, **ya; // for input vectors xa[0...xlen-1][0..2] and // ya[0...ylen-1][0..2], in general, // ya is regarded as native structure // --> superpose xa onto ya vector resi_vec1; // residue index for chain1 vector resi_vec2; // residue index for chain2 int read_resi=byresi_opt; // whether to read residue index if (byresi_opt==0 && o_opt) read_resi=2; /* loop over file names */ for (i=0;i0) make_sec(seqx,xa, xlen, secx,atom_opt); else make_sec(xa, xlen, secx); // secondary structure assignment for (j=(dir_opt.size()>0)*(i+1);j0) make_sec(seqy, ya, ylen, secy, atom_opt); else make_sec(ya, ylen, secy); if (byresi_opt) extract_aln_from_resi(sequence, seqx,seqy,resi_vec1,resi_vec2,byresi_opt); /* declare variable specific to this pair of TMalign */ double t0[3], u0[3][3]; double TM1, TM2; double TM3, TM4, TM5; // for a_opt, u_opt, d_opt double d0_0, TM_0; double d0A, d0B, d0u, d0a; double d0_out=5.0; string seqM, seqxA, seqyA;// for output alignment double rmsd0 = 0.0; int L_ali; // Aligned length in standard_TMscore double Liden=0; double TM_ali, rmsd_ali; // TMscore and rmsd in standard_TMscore int n_ali=0; int n_ali8=0; bool force_fast_opt=(getmin(xlen,ylen)>1500)?true:fast_opt; vector do_vec; /* entry function for structure alignment */ if (cp_opt) CPalign_main( xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, force_fast_opt, mol_vec1[chain_i]+mol_vec2[chain_j],TMcut); else if (se_opt) { int *invmap = new int[ylen+1]; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, mol_vec1[chain_i]+mol_vec2[chain_j], outfmt_opt, invmap); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r21) { yname.clear(); for (chain_j=0;chain_j &sequence, const double d0_scale, const bool m_opt, const int o_opt, const int a_opt, const bool d_opt, const bool full_opt, const double TMcut, const int infmt1_opt, const int infmt2_opt, const int ter_opt, const int split_opt, const int outfmt_opt, bool fast_opt, const int mirror_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir1_opt, const string &dir2_opt, const vector &chain2parse1, const vector &chain2parse2, const vector &model2parse1, const vector &model2parse2, const vector &chain1_list, const vector &chain2_list, const int byresi_opt,const string&chainmapfile, const bool se_opt) { /* declare previously global variables */ vector > > xa_vec; // structure of complex1 vector > > ya_vec; // structure of complex2 vector >seqx_vec; // sequence of complex1 vector >seqy_vec; // sequence of complex2 vector >secx_vec; // secondary structure of complex1 vector >secy_vec; // secondary structure of complex2 vector mol_vec1; // molecule type of complex1, RNA if >0 vector mol_vec2; // molecule type of complex2, RNA if >0 vector chainID_list1; // list of chainID1 vector chainID_list2; // list of chainID2 vector xlen_vec; // length of complex1 vector ylen_vec; // length of complex2 int i,j; // chain index int xlen, ylen; // chain length double **xa, **ya; // structure of single chain char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure int xlen_aa,ylen_aa; // total length of protein int xlen_na,ylen_na; // total length of RNA/DNA vector resi_vec1; // residue index for chain1 vector resi_vec2; // residue index for chain2 /* parse complex */ parse_chain_list(chain1_list, xa_vec, seqx_vec, secx_vec, mol_vec1, xlen_vec, chainID_list1, ter_opt, split_opt, mol_opt, infmt1_opt, atom_opt, autojustify, mirror_opt, het_opt, xlen_aa, xlen_na, o_opt, resi_vec1, chain2parse1, model2parse1); if (xa_vec.size()==0) PrintErrorAndQuit("ERROR! 0 chain in complex 1"); parse_chain_list(chain2_list, ya_vec, seqy_vec, secy_vec, mol_vec2, ylen_vec, chainID_list2, ter_opt, split_opt, mol_opt, infmt2_opt, atom_opt, autojustify, 0, het_opt, ylen_aa, ylen_na, o_opt, resi_vec2, chain2parse2, model2parse2); if (ya_vec.size()==0) PrintErrorAndQuit("ERROR! 0 chain in complex 2"); int len_aa=getmin(xlen_aa,ylen_aa); int len_na=getmin(xlen_na,ylen_na); if (a_opt) { len_aa=(xlen_aa+ylen_aa)/2; len_na=(xlen_na+ylen_na)/2; } int i_opt=0; if (byresi_opt) i_opt=3; map chainmap; if (chainmapfile.size()) { string line; int chainidx1,chainidx2; vector line_vec; ifstream fin; bool fromStdin=(chainmapfile=="-"); if (!fromStdin) fin.open(chainmapfile.c_str()); while (fromStdin?cin.good():fin.good()) { if (fromStdin) getline(cin,line); else getline(fin,line); if (line.size()==0 || line[0]=='#') continue; split(line,line_vec,'\t'); if (line_vec.size()==2) { chainidx1=-1; chainidx2=-1; for (i=0;i=0 && chainidx2>=0) { if (chainmap.count(chainidx1)) cerr<<"ERROR! "< do_vec; if (byresi_opt) extract_aln_from_resi(sequence, seqx,seqy,resi_vec1,resi_vec2,byresi_opt); /* entry function for structure alignment */ if (se_opt) { int *invmap = new int[ylen+1]; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, 0, d0_scale, i_opt, a_opt, false, d_opt, mol_vec1[0]+mol_vec2[0], outfmt_opt, invmap); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r2 > >().swap(xa_vec); // structure of complex1 vector > >().swap(ya_vec); // structure of complex2 vector >().swap(seqx_vec); // sequence of complex1 vector >().swap(seqy_vec); // sequence of complex2 vector >().swap(secx_vec); // secondary structure of complex1 vector >().swap(secy_vec); // secondary structure of complex2 mol_vec1.clear(); // molecule type of complex1, RNA if >0 mol_vec2.clear(); // molecule type of complex2, RNA if >0 chainID_list1.clear(); // list of chainID1 chainID_list2.clear(); // list of chainID2 xlen_vec.clear(); // length of complex1 ylen_vec.clear(); // length of complex2 return 0; } /* declare TM-score tables */ int chain1_num=xa_vec.size(); int chain2_num=ya_vec.size(); vector tmp_str_vec(chain2_num,""); double **TMave_mat; double **ut_mat; // rotation matrices for all-against-all alignment int ui,uj,ut_idx; NewArray(&TMave_mat,chain1_num,chain2_num); NewArray(&ut_mat,chain1_num*chain2_num,4*3); vector >seqxA_mat(chain1_num,tmp_str_vec); vector > seqM_mat(chain1_num,tmp_str_vec); vector >seqyA_mat(chain1_num,tmp_str_vec); double maxTMmono=-1; int maxTMmono_i,maxTMmono_j; /* get all-against-all alignment */ if (len_aa+len_na>500) fast_opt=true; for (i=0;i do_vec; int Lnorm_tmp=len_aa; if (mol_vec1[i]+mol_vec2[j]>0) Lnorm_tmp=len_na; if (byresi_opt) { int total_aln=extract_aln_from_resi(sequence, seqx,seqy, resi_vec1,resi_vec2,xlen_vec,ylen_vec, i, j, byresi_opt); seqxA_mat[i][j]=sequence[0]; seqyA_mat[i][j]=sequence[1]; if (total_aln>xlen+ylen-3) { for (ui=0;ui<3;ui++) for (uj=0;uj<3;uj++) ut_mat[ut_idx][ui*3+uj]=(ui==uj)?1:0; for (uj=0;uj<3;uj++) ut_mat[ut_idx][9+uj]=0; TMave_mat[i][j]=0; seqM.clear(); seqxA.clear(); seqyA.clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); continue; } } /* entry function for structure alignment */ if (se_opt) { int *invmap = new int[ylen+1]; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_tmp, d0_scale, i_opt, false, true, false, mol_vec1[i]+mol_vec2[j], outfmt_opt, invmap); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r2maxTMmono) { maxTMmono=TMave_mat[i][j]; maxTMmono_i=i; maxTMmono_j=j; } /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); } /* calculate initial chain-chain assignment */ int *assign1_list; // value is index of assigned chain2 int *assign2_list; // value is index of assigned chain1 assign1_list=new int[chain1_num]; assign2_list=new int[chain2_num]; double total_score=enhanced_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num); if (total_score<=0) PrintErrorAndQuit("ERROR! No assignable chain"); /* refine alignment for large oligomers */ int aln_chain_num=count_assign_pair(assign1_list,chain1_num); bool is_oligomer=(aln_chain_num>=3); if (aln_chain_num==2 && chainmap.size()==0 && !se_opt) // dimer alignment { int na_chain_num1,na_chain_num2,aa_chain_num1,aa_chain_num2; count_na_aa_chain_num(na_chain_num1,aa_chain_num1,mol_vec1); count_na_aa_chain_num(na_chain_num2,aa_chain_num2,mol_vec2); /* align protein-RNA hybrid dimer to another hybrid dimer */ if (na_chain_num1==1 && na_chain_num2==1 && aa_chain_num1==1 && aa_chain_num2==1) is_oligomer=false; /* align pure protein dimer or pure RNA dimer */ else if ((getmin(na_chain_num1,na_chain_num2)==0 && aa_chain_num1==2 && aa_chain_num2==2) || (getmin(aa_chain_num1,aa_chain_num2)==0 && na_chain_num1==2 && na_chain_num2==2)) { adjust_dimer_assignment(xa_vec,ya_vec,xlen_vec,ylen_vec,mol_vec1, mol_vec2,assign1_list,assign2_list,seqxA_mat,seqyA_mat); is_oligomer=false; // cannot refiner further } else is_oligomer=true; /* align oligomers to dimer */ } if ((aln_chain_num>=3 || is_oligomer) && chainmap.size()==0 && !se_opt) // oligomer alignment { /* extract centroid coordinates */ double **xcentroids; double **ycentroids; NewArray(&xcentroids, chain1_num, 3); NewArray(&ycentroids, chain2_num, 3); double d0MM=getmin( calculate_centroids(xa_vec, chain1_num, xcentroids), calculate_centroids(ya_vec, chain2_num, ycentroids)); /* refine enhanced greedy search with centroid superposition */ //double het_deg=check_heterooligomer(TMave_mat, chain1_num, chain2_num); homo_refined_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num, xcentroids, ycentroids, d0MM, len_aa+len_na, ut_mat); hetero_refined_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num, xcentroids, ycentroids, d0MM, len_aa+len_na); /* clean up */ DeleteArray(&xcentroids, chain1_num); DeleteArray(&ycentroids, chain2_num); } /* store initial assignment */ int init_pair_num=count_assign_pair(assign1_list,chain1_num); int *assign1_init, *assign2_init; assign1_init=new int[chain1_num]; assign2_init=new int[chain2_num]; double **TMave_init; NewArray(&TMave_init,chain1_num,chain2_num); vector >seqxA_init(chain1_num,tmp_str_vec); vector >seqyA_init(chain1_num,tmp_str_vec); vector sequence_init; copy_chain_assign_data(chain1_num, chain2_num, sequence_init, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat, seqxA_init, seqyA_init, assign1_init, assign2_init, TMave_init); /* perform iterative alignment */ double max_total_score=0; // ignore old total_score because previous // score was from monomeric chain superpositions int max_iter=5-(int)((len_aa+len_na)/200); if (max_iter<2) max_iter=2; //if (byresi_opt==0) if (!se_opt) MMalign_iter(max_total_score, max_iter, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_mat, seqxA_mat, seqyA_mat, assign1_list, assign2_list, sequence, d0_scale, fast_opt, chainmap, byresi_opt); if (byresi_opt && aln_chain_num>=4 && is_oligomer && chainmap.size()==0 && !se_opt) // oligomer alignment { MMalign_final(xname.substr(dir1_opt.size()), yname.substr(dir2_opt.size()), chainID_list1, chainID_list2, fname_super, fname_lign, fname_matrix, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_mat, seqxA_mat, seqM_mat, seqyA_mat, assign1_list, assign2_list, sequence, d0_scale, 1, 0, 5, ter_opt, split_opt, 0, 0, true, true, mirror_opt, resi_vec1, resi_vec2); /* extract centroid coordinates */ double **xcentroids; double **ycentroids; NewArray(&xcentroids, chain1_num, 3); NewArray(&ycentroids, chain2_num, 3); double d0MM=getmin( calculate_centroids(xa_vec, chain1_num, xcentroids), calculate_centroids(ya_vec, chain2_num, ycentroids)); /* refine enhanced greedy search with centroid superposition */ //double het_deg=check_heterooligomer(TMave_mat, chain1_num, chain2_num); homo_refined_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num, xcentroids, ycentroids, d0MM, len_aa+len_na, ut_mat); hetero_refined_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num, xcentroids, ycentroids, d0MM, len_aa+len_na); /* clean up */ DeleteArray(&xcentroids, chain1_num); DeleteArray(&ycentroids, chain2_num); } /* sometime MMalign_iter is even worse than monomer alignment */ if (byresi_opt==0 && max_total_score=init_pair_num) copy_chain_assign_data( chain1_num, chain2_num, sequence_init, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat, seqxA_init, seqyA_init, assign1_init, assign2_init, TMave_init); double max_total_score_cross=max_total_score; if (byresi_opt==0 && len_aa+len_na<10000) { MMalign_dimer(max_total_score_cross, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_init, seqxA_init, seqyA_init, assign1_init, assign2_init, sequence_init, d0_scale, fast_opt); if (max_total_score_cross>max_total_score) { max_total_score=max_total_score_cross; copy_chain_assign_data(chain1_num, chain2_num, sequence, seqxA_init, seqyA_init, assign1_init, assign2_init, TMave_init, seqxA_mat, seqyA_mat, assign1_list, assign2_list, TMave_mat); } } /* final alignment */ if (outfmt_opt==0) print_version(); if (se_opt) MMalign_se_final(xname.substr(dir1_opt.size()), yname.substr(dir2_opt.size()), chainID_list1, chainID_list2, fname_super, fname_lign, fname_matrix, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_mat, seqxA_mat, seqM_mat, seqyA_mat, assign1_list, assign2_list, sequence, d0_scale, m_opt, o_opt, outfmt_opt, ter_opt, split_opt, a_opt, d_opt, fast_opt, full_opt, mirror_opt, resi_vec1, resi_vec2); else MMalign_final(xname.substr(dir1_opt.size()), yname.substr(dir2_opt.size()), chainID_list1, chainID_list2, fname_super, fname_lign, fname_matrix, xa_vec, ya_vec, seqx_vec, seqy_vec, secx_vec, secy_vec, mol_vec1, mol_vec2, xlen_vec, ylen_vec, xa, ya, seqx, seqy, secx, secy, len_aa, len_na, chain1_num, chain2_num, TMave_mat, seqxA_mat, seqM_mat, seqyA_mat, assign1_list, assign2_list, sequence, d0_scale, m_opt, o_opt, outfmt_opt, ter_opt, split_opt, a_opt, d_opt, fast_opt, full_opt, mirror_opt, resi_vec1, resi_vec2); /* clean up everything */ delete [] assign1_list; delete [] assign2_list; DeleteArray(&TMave_mat,chain1_num); DeleteArray(&ut_mat, chain1_num*chain2_num); vector >().swap(seqxA_mat); vector >().swap(seqM_mat); vector >().swap(seqyA_mat); vector().swap(tmp_str_vec); delete [] assign1_init; delete [] assign2_init; DeleteArray(&TMave_init,chain1_num); vector >().swap(seqxA_init); vector >().swap(seqyA_init); vector > >().swap(xa_vec); // structure of complex1 vector > >().swap(ya_vec); // structure of complex2 vector >().swap(seqx_vec); // sequence of complex1 vector >().swap(seqy_vec); // sequence of complex2 vector >().swap(secx_vec); // secondary structure of complex1 vector >().swap(secy_vec); // secondary structure of complex2 mol_vec1.clear(); // molecule type of complex1, RNA if >0 mol_vec2.clear(); // molecule type of complex2, RNA if >0 vector().swap(chainID_list1); // list of chainID1 vector().swap(chainID_list2); // list of chainID2 xlen_vec.clear(); // length of complex1 ylen_vec.clear(); // length of complex2 vector ().swap(resi_vec1); // residue index for chain1 vector ().swap(resi_vec2); // residue index for chain2 map ().swap(chainmap); return 1; } /* alignment individual chains to a complex. */ int MMdock(const string &xname, const string &yname, const string &fname_super, const string &fname_matrix, vector &sequence, const double Lnorm_ass, const double d0_scale, const bool m_opt, const int o_opt, const int a_opt, const bool u_opt, const bool d_opt, const double TMcut, const int infmt1_opt, const int infmt2_opt, const int ter_opt, const int split_opt, const int outfmt_opt, bool fast_opt, const int mirror_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir1_opt, const string &dir2_opt, const vector &chain2parse1, const vector &chain2parse2, const vector &model2parse1, const vector &model2parse2, const vector &chain1_list, const vector &chain2_list, const bool do_opt) { /* declare previously global variables */ vector > > xa_vec; // structure of complex1 vector > > ya_vec; // structure of complex2 vector >seqx_vec; // sequence of complex1 vector >seqy_vec; // sequence of complex2 vector >secx_vec; // secondary structure of complex1 vector >secy_vec; // secondary structure of complex2 vector mol_vec1; // molecule type of complex1, RNA if >0 vector mol_vec2; // molecule type of complex2, RNA if >0 vector chainID_list1; // list of chainID1 vector chainID_list2; // list of chainID2 vector xlen_vec; // length of complex1 vector ylen_vec; // length of complex2 int i,j; // chain index int xlen, ylen; // chain length double **xa, **ya; // structure of single chain char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure int xlen_aa,ylen_aa; // total length of protein int xlen_na,ylen_na; // total length of RNA/DNA vector resi_vec1; // residue index for chain1 vector resi_vec2; // residue index for chain2 /* parse complex */ parse_chain_list(chain1_list, xa_vec, seqx_vec, secx_vec, mol_vec1, xlen_vec, chainID_list1, ter_opt, split_opt, mol_opt, infmt1_opt, atom_opt, autojustify, mirror_opt, het_opt, xlen_aa, xlen_na, o_opt, resi_vec1, chain2parse1, model2parse1); if (xa_vec.size()==0) PrintErrorAndQuit("ERROR! 0 individual chain"); parse_chain_list(chain2_list, ya_vec, seqy_vec, secy_vec, mol_vec2, ylen_vec, chainID_list2, ter_opt, split_opt, mol_opt, infmt2_opt, atom_opt, autojustify, 0, het_opt, ylen_aa, ylen_na, o_opt, resi_vec2, chain2parse2, model2parse2); if (xa_vec.size()>ya_vec.size()) PrintErrorAndQuit( "ERROR! more individual chains to align than number of chains in complex template"); int len_aa=getmin(xlen_aa,ylen_aa); int len_na=getmin(xlen_na,ylen_na); if (a_opt) { len_aa=(xlen_aa+ylen_aa)/2; len_na=(xlen_na+ylen_na)/2; } /* perform monomer alignment if there is only one chain */ if (xa_vec.size()==1 && ya_vec.size()==1) { xlen = xlen_vec[0]; ylen = ylen_vec[0]; seqx = new char[xlen+1]; seqy = new char[ylen+1]; secx = new char[xlen+1]; secy = new char[ylen+1]; NewArray(&xa, xlen, 3); NewArray(&ya, ylen, 3); copy_chain_data(xa_vec[0],seqx_vec[0],secx_vec[0], xlen,xa,seqx,secx); copy_chain_data(ya_vec[0],seqy_vec[0],secy_vec[0], ylen,ya,seqy,secy); /* declare variable specific to this pair of TMalign */ double t0[3], u0[3][3]; double TM1, TM2; double TM3, TM4, TM5; // for a_opt, u_opt, d_opt double d0_0, TM_0; double d0A, d0B, d0u, d0a; double d0_out=5.0; string seqM, seqxA, seqyA;// for output alignment double rmsd0 = 0.0; int L_ali; // Aligned length in standard_TMscore double Liden=0; double TM_ali, rmsd_ali; // TMscore and rmsd in standard_TMscore int n_ali=0; int n_ali8=0; vector do_vec; /* entry function for structure alignment */ TMalign_main(xa, ya, seqx, seqy, secx, secy, t0, u0, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 0, a_opt, u_opt, d_opt, fast_opt, mol_vec1[0]+mol_vec2[0],TMcut); /* print result */ output_results( xname.substr(dir1_opt.size()), yname.substr(dir2_opt.size()), chainID_list1[0], chainID_list2[0], xlen, ylen, t0, u0, TM1, TM2, TM3, TM4, TM5, rmsd0, d0_out, seqM.c_str(), seqxA.c_str(), seqyA.c_str(), Liden, n_ali8, L_ali, TM_ali, rmsd_ali, TM_0, d0_0, d0A, d0B, Lnorm_ass, d0_scale, d0a, d0u, (m_opt?fname_matrix:"").c_str(), (outfmt_opt==2?outfmt_opt:3), ter_opt, true, split_opt, o_opt, fname_super, 0, a_opt, false, d_opt, mirror_opt, resi_vec1, resi_vec2); if (outfmt_opt==2) printf("%s%s\t%s%s\t%.4f\n", xname.substr(dir1_opt.size()).c_str(), chainID_list1[0].c_str(), yname.substr(dir2_opt.size()).c_str(), chainID_list2[0].c_str(), sqrt((TM1*TM1+TM2*TM2)/2)); /* clean up */ seqM.clear(); seqxA.clear(); seqyA.clear(); delete[]seqx; delete[]seqy; delete[]secx; delete[]secy; DeleteArray(&xa,xlen); DeleteArray(&ya,ylen); do_vec.clear(); vector > >().swap(xa_vec); // structure of complex1 vector > >().swap(ya_vec); // structure of complex2 vector >().swap(seqx_vec); // sequence of complex1 vector >().swap(seqy_vec); // sequence of complex2 vector >().swap(secx_vec); // secondary structure of complex1 vector >().swap(secy_vec); // secondary structure of complex2 mol_vec1.clear(); // molecule type of complex1, RNA if >0 mol_vec2.clear(); // molecule type of complex2, RNA if >0 chainID_list1.clear(); // list of chainID1 chainID_list2.clear(); // list of chainID2 xlen_vec.clear(); // length of complex1 ylen_vec.clear(); // length of complex2 return 0; } /* declare TM-score tables */ int chain1_num=xa_vec.size(); int chain2_num=ya_vec.size(); vector tmp_str_vec(chain2_num,""); double **TMave_mat; NewArray(&TMave_mat,chain1_num,chain2_num); vector >seqxA_mat(chain1_num,tmp_str_vec); vector > seqM_mat(chain1_num,tmp_str_vec); vector >seqyA_mat(chain1_num,tmp_str_vec); /* trimComplex */ vector > > ya_trim_vec; // structure of complex2 vector >seqy_trim_vec; // sequence of complex2 vector >secy_trim_vec; // secondary structure of complex2 vector ylen_trim_vec; // length of complex2 int Lchain_aa_max1=0; int Lchain_na_max1=0; for (i=0;i0 && xlen>Lchain_na_max1) Lchain_na_max1=xlen; else if (mol_vec1[i]<=0 && xlen>Lchain_aa_max1) Lchain_aa_max1=xlen; } int trim_chain_count=trimComplex(ya_trim_vec,seqy_trim_vec, secy_trim_vec,ylen_trim_vec,ya_vec,seqy_vec,secy_vec,ylen_vec, mol_vec2,Lchain_aa_max1,Lchain_na_max1); int ylen_trim; // chain length double **ya_trim; // structure of single chain char *seqy_trim; // for the protein sequence char *secy_trim; // for the secondary structure double **xt; /* get all-against-all alignment */ if (len_aa+len_na>500) fast_opt=true; for (i=0;i do_vec; int Lnorm_tmp=len_aa; if (mol_vec1[i]+mol_vec2[j]>0) Lnorm_tmp=len_na; /* entry function for structure alignment */ if (trim_chain_count && ylen_trim_vec[j] > >().swap(ya_trim_vec); vector >().swap(seqy_trim_vec); vector >().swap(secy_trim_vec); vector ().swap(ylen_trim_vec); /* calculate initial chain-chain assignment */ int *assign1_list; // value is index of assigned chain2 int *assign2_list; // value is index of assigned chain1 assign1_list=new int[chain1_num]; assign2_list=new int[chain2_num]; enhanced_greedy_search(TMave_mat, assign1_list, assign2_list, chain1_num, chain2_num); /* final alignment */ if (outfmt_opt==0) print_version(); double **ut_mat; // rotation matrices for all-against-all alignment NewArray(&ut_mat,chain1_num,4*3); int ui,uj; vectorxname_vec; vectoryname_vec; vectorTM_vec; for (i=0;i do_vec; int c; for (c=0; c().swap(TM_vec); vector().swap(xname_vec); vector().swap(yname_vec); delete [] assign1_list; delete [] assign2_list; DeleteArray(&TMave_mat,chain1_num); DeleteArray(&ut_mat, chain1_num); vector >().swap(seqxA_mat); vector >().swap(seqM_mat); vector >().swap(seqyA_mat); vector().swap(tmp_str_vec); vector > >().swap(xa_vec); // structure of complex1 vector > >().swap(ya_vec); // structure of complex2 vector >().swap(seqx_vec); // sequence of complex1 vector >().swap(seqy_vec); // sequence of complex2 vector >().swap(secx_vec); // secondary structure of complex1 vector >().swap(secy_vec); // secondary structure of complex2 mol_vec1.clear(); // molecule type of complex1, RNA if >0 mol_vec2.clear(); // molecule type of complex2, RNA if >0 vector().swap(chainID_list1); // list of chainID1 vector().swap(chainID_list2); // list of chainID2 xlen_vec.clear(); // length of complex1 ylen_vec.clear(); // length of complex2 return 1; } int mTMalign(string &xname, string &yname, const string &fname_super, const string &fname_matrix, vector &sequence, double Lnorm_ass, const double d0_scale, const bool m_opt, const int i_opt, const int o_opt, const int a_opt, bool u_opt, const bool d_opt, const bool full_opt, const double TMcut, const int infmt_opt, const int ter_opt, const int split_opt, const int outfmt_opt, bool fast_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir_opt, const int byresi_opt, const vector &chain_list, const vector &chain2parse, const vector &model2parse, const bool se_opt) { /* declare previously global variables */ vector > >a_vec; // atomic structure vector > >ua_vec; // unchanged atomic structure vector >seq_vec; // sequence of complex vector >sec_vec; // secondary structure of complex vector mol_vec; // molecule type of complex1, RNA if >0 vector chainID_list; // list of chainID vector len_vec; // length of complex int i,j; // chain index int xlen, ylen; // chain length double **xa, **ya; // structure of single chain char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure int len_aa,len_na; // total length of protein and RNA/DNA vector resi_vec; // residue index for chain /* parse chain list */ parse_chain_list(chain_list, a_vec, seq_vec, sec_vec, mol_vec, len_vec, chainID_list, ter_opt, split_opt, mol_opt, infmt_opt, atom_opt, autojustify, false, het_opt, len_aa, len_na, o_opt, resi_vec, chain2parse, model2parse); int chain_num=a_vec.size(); if (chain_num<=1) PrintErrorAndQuit("ERROR! <2 chains for multiple alignment"); if (m_opt||o_opt) for (i=0;ixlen) xlen=len_vec[i]; total_len+=len_vec[i]; mol_type+=mol_vec[i]; } if (!u_opt) Lnorm_ass=total_len/chain_num; u_opt=true; total_len-=xlen; if (total_len>750) fast_opt=true; /* get all-against-all alignment */ double **TMave_mat; NewArray(&TMave_mat,chain_num,chain_num); vector tmp_str_vec(chain_num,""); vector >seqxA_mat(chain_num,tmp_str_vec); vector >seqyA_mat(chain_num,tmp_str_vec); for (i=0;i do_vec; /* entry function for structure alignment */ if (se_opt) { int *invmap = new int[ylen+1]; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 0, false, u_opt, false, mol_type, outfmt_opt, invmap); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r2xname_vec; for (i=0;iyname_vec; double *TMave_list; TMave_list = new double[chain_num]; int *assign_list; assign_list=new int[chain_num]; vector msa(ylen,""); // row is position along msa; column is sequence int compare_num; double TM1_total, TM2_total; double TM3_total, TM4_total, TM5_total; // for a_opt, u_opt, d_opt double d0_0_total, TM_0_total; double d0A_total, d0B_total, d0u_total, d0a_total; double d0_out_total; double rmsd0_total; int L_ali_total; // Aligned length in standard_TMscore double Liden_total; double TM_ali_total, rmsd_ali_total; // TMscore and rmsd in standard_TMscore int n_ali_total; int n_ali8_total; int xlen_total, ylen_total; double TM4_total_max=0; int max_iter=5-(int)(total_len/200); if (max_iter<2) max_iter=2; int iter=0; vector TM_vec(chain_num,0); vector d0_vec(chain_num,0); vector seqID_vec(chain_num,0); vector > TM_mat(chain_num,TM_vec); vector > d0_mat(chain_num,d0_vec); vector > seqID_mat(chain_num,seqID_vec); for (iter=0; iter do_vec; /* entry function for structure alignment */ if (se_opt) { int *invmap = new int[ylen+1]; u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 2, a_opt, u_opt, d_opt, mol_type, outfmt_opt, invmap); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r2 msa_ext; // row is position along msa; column is sequence for (r=0;r do_vec; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, 0, a_opt, u_opt, d_opt, mol_type, 1, invmap); int rx=0,ry=0; ylen_ext=seqxA.size(); NewArray(&ya_ext, ylen_ext, 3); // structure of single chain seqy_ext= new char[ylen_ext+1]; // for the protein sequence secy_ext= new char[ylen_ext+1]; // for the secondary structure string tmp_gap=""; for (r=0;r().swap(msa_ext); vector >().swap(TM_pair_vec); for (i=0; i do_vec; se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, true, a_opt, u_opt, d_opt, mol_type, 1, invmap); if (xlen<=ylen) { xlen_total+=xlen; ylen_total+=ylen; TM1_total+=TM1; TM2_total+=TM2; d0A_total+=d0A; d0B_total+=d0B; } else { xlen_total+=ylen; ylen_total+=xlen; TM1_total+=TM2; TM2_total+=TM1; d0A_total+=d0B; d0B_total+=d0A; } TM_mat[i][j]=TM2; TM_mat[j][i]=TM1; d0_mat[i][j]=d0B; d0_mat[j][i]=d0A; seqID_mat[i][j]=1.*Liden/xlen; seqID_mat[j][i]=1.*Liden/ylen; TM3_total+=TM3; TM4_total+=TM4; TM5_total+=TM5; d0_0_total+=d0_0; TM_0_total+=TM_0; d0u_total+=d0u; d0_out_total+=d0_out; rmsd0_total+=rmsd0; L_ali_total+=L_ali; // Aligned length in standard_TMscore Liden_total+=Liden; TM_ali_total+=TM_ali; rmsd_ali_total+=rmsd_ali; // TMscore and rmsd in standard_TMscore n_ali_total+=n_ali; n_ali8_total+=n_ali8; /* clean up */ delete[]invmap; seqM.clear(); seqxA.clear(); seqyA.clear(); delete[]seqy; delete[]secy; DeleteArray(&ya,ylen); do_vec.clear(); } delete[]seqx; delete[]secx; DeleteArray(&xa,xlen); } if (TM4_total<=TM4_total_max) break; TM4_total_max=TM4_total; } for (i=0;i"< > >().swap(ua_vec); if (m_opt) { assign_list[repr_idx]=-1; output_dock_rotation_matrix(fname_matrix.c_str(), xname_vec,yname_vec, ut_mat, assign_list); } if (o_opt) output_dock(chain_list, ter_opt, split_opt, infmt_opt, atom_opt, false, ut_mat, fname_super); DeleteArray(&ut_mat,chain_num); } /* clean up */ vector().swap(msa); vector().swap(tmp_str_vec); vector >().swap(seqxA_mat); vector >().swap(seqyA_mat); vector().swap(xname_vec); vector().swap(yname_vec); delete[]TMave_list; DeleteArray(&TMave_mat,chain_num); vector > >().swap(a_vec); // structure of complex vector >().swap(seq_vec); // sequence of complex vector >().swap(sec_vec); // secondary structure of complex vector().swap(mol_vec); // molecule type of complex1, RNA if >0 vector().swap(chainID_list); // list of chainID vector().swap(len_vec); // length of complex vector().swap(TM_vec); vector().swap(d0_vec); vector().swap(seqID_vec); vector >().swap(TM_mat); vector >().swap(d0_mat); vector >().swap(seqID_mat); return 1; } /* sequence order independent alignment */ int SOIalign(string &xname, string &yname, const string &fname_super, const string &fname_lign, const string &fname_matrix, vector &sequence, const double Lnorm_ass, const double d0_scale, const bool m_opt, const int i_opt, const int o_opt, const int a_opt, const bool u_opt, const bool d_opt, const double TMcut, const int infmt1_opt, const int infmt2_opt, const int ter_opt, const int split_opt, const int outfmt_opt, const bool fast_opt, const int cp_opt, const int mirror_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir_opt, const string &dirpair_opt, const string &dir1_opt, const string &dir2_opt, const vector &chain2parse1, const vector &chain2parse2, const vector &model2parse1, const vector &model2parse2, const vector &chain1_list, const vector &chain2_list, const bool se_opt, const int closeK_opt, const int mm_opt) { /* declare previously global variables */ vector >PDB_lines1; // text of chain1 vector >PDB_lines2; // text of chain2 vector mol_vec1; // molecule type of chain1, RNA if >0 vector mol_vec2; // molecule type of chain2, RNA if >0 vector chainID_list1; // list of chainID1 vector chainID_list2; // list of chainID2 int i,j; // file index int chain_i,chain_j; // chain index int r; // residue index int xlen, ylen; // chain length int xchainnum,ychainnum;// number of chains in a PDB file char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure int **secx_bond; // boundary of secondary structure int **secy_bond; // boundary of secondary structure double **xa, **ya; // for input vectors xa[0...xlen-1][0..2] and // ya[0...ylen-1][0..2], in general, // ya is regarded as native structure // --> superpose xa onto ya double **xk, **yk; // k closest residues vector resi_vec1; // residue index for chain1 vector resi_vec2; // residue index for chain2 int read_resi=0; // whether to read residue index if (o_opt) read_resi=2; /* loop over file names */ for (i=0;i=3) NewArray(&xk, xlen*closeK_opt, 3); seqx = new char[xlen + 1]; secx = new char[xlen + 1]; xlen = read_PDB(PDB_lines1[chain_i], xa, seqx, resi_vec1, read_resi); if (mirror_opt) for (r=0;r0) make_sec(seqx,xa, xlen, secx,atom_opt); else make_sec(xa, xlen, secx); // secondary structure assignment if (closeK_opt>=3) getCloseK(xa, xlen, closeK_opt, xk); if (mm_opt==6) { NewArray(&secx_bond, xlen, 2); assign_sec_bond(secx_bond, secx, xlen); } for (j=(dir_opt.size()>0)*(i+1);j=3) NewArray(&yk, ylen*closeK_opt, 3); seqy = new char[ylen + 1]; secy = new char[ylen + 1]; ylen = read_PDB(PDB_lines2[chain_j], ya, seqy, resi_vec2, read_resi); if (mol_vec2[chain_j]>0) make_sec(seqy, ya, ylen, secy, atom_opt); else make_sec(ya, ylen, secy); if (closeK_opt>=3) getCloseK(ya, ylen, closeK_opt, yk); if (mm_opt==6) { NewArray(&secy_bond, ylen, 2); assign_sec_bond(secy_bond, secy, ylen); } /* declare variable specific to this pair of TMalign */ double t0[3], u0[3][3]; double TM1, TM2; double TM3, TM4, TM5; // for a_opt, u_opt, d_opt double d0_0, TM_0; double d0A, d0B, d0u, d0a; double d0_out=5.0; string seqM, seqxA, seqyA;// for output alignment double rmsd0 = 0.0; int L_ali; // Aligned length in standard_TMscore double Liden=0; double TM_ali, rmsd_ali; // TMscore and rmsd in standard_TMscore int n_ali=0; int n_ali8=0; bool force_fast_opt=(getmin(xlen,ylen)>1500)?true:fast_opt; int *invmap = new int[ylen+1]; double *dist_list = new double[ylen+1]; /* entry function for structure alignment */ if (se_opt) { u0[0][0]=u0[1][1]=u0[2][2]=1; u0[0][1]= u0[0][2]= u0[1][0]= u0[1][2]= u0[2][0]= u0[2][1]= t0[0] =t0[1] =t0[2] =0; soi_se_main(xa, ya, seqx, seqy, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, mol_vec1[chain_i]+mol_vec2[chain_j], outfmt_opt, invmap, dist_list, secx_bond, secy_bond, mm_opt); if (outfmt_opt>=2) { Liden=L_ali=0; int r1,r2; for (r2=0;r2=3) DeleteArray(&yk, ylen*closeK_opt); delete [] seqy; delete [] secy; resi_vec2.clear(); if (mm_opt==6) DeleteArray(&secy_bond, ylen); } // chain_j if (chain2_list.size()>1) { yname.clear(); for (chain_j=0;chain_j=3) DeleteArray(&xk, xlen*closeK_opt); delete [] seqx; delete [] secx; resi_vec1.clear(); if (mm_opt==6) DeleteArray(&secx_bond, xlen); } // chain_i xname.clear(); PDB_lines1.clear(); chainID_list1.clear(); mol_vec1.clear(); } // i if (chain2_list.size()==1) { yname.clear(); for (chain_j=0;chain_j &sequence, const double Lnorm_ass, const double d0_scale, const bool m_opt, const int i_opt, const int o_opt, const int a_opt, const bool u_opt, const bool d_opt, const double TMcut, const int infmt1_opt, const int infmt2_opt, const int ter_opt, const int split_opt, const int outfmt_opt, const bool fast_opt, const int mirror_opt, const int het_opt, const string &atom_opt, const bool autojustify, const string &mol_opt, const string &dir_opt, const string &dirpair_opt, const string &dir1_opt, const string &dir2_opt, const vector &chain2parse1, const vector &chain2parse2, const vector &model2parse1, const vector &model2parse2, const int byresi_opt, const vector &chain1_list, const vector &chain2_list, const int hinge_opt) { /* declare previously global variables */ vector >PDB_lines1; // text of chain1 vector >PDB_lines2; // text of chain2 vector mol_vec1; // molecule type of chain1, RNA if >0 vector mol_vec2; // molecule type of chain2, RNA if >0 vector chainID_list1; // list of chainID1 vector chainID_list2; // list of chainID2 int i,j; // file index int chain_i,chain_j; // chain index int r; // residue index int xlen, ylen; // chain length int xchainnum,ychainnum;// number of chains in a PDB file char *seqx, *seqy; // for the protein sequence char *secx, *secy; // for the secondary structure double **xa, **ya; // for input vectors xa[0...xlen-1][0..2] and // ya[0...ylen-1][0..2], in general, // ya is regarded as native structure // --> superpose xa onto ya vector resi_vec1; // residue index for chain1 vector resi_vec2; // residue index for chain2 int read_resi=byresi_opt; // whether to read residue index if (byresi_opt==0 && o_opt) read_resi=2; /* loop over file names */ for (i=0;i0) make_sec(seqx,xa, xlen, secx,atom_opt); else make_sec(xa, xlen, secx); // secondary structure assignment for (j=(dir_opt.size()>0)*(i+1);j0) make_sec(seqy, ya, ylen, secy, atom_opt); else make_sec(ya, ylen, secy); if (byresi_opt) extract_aln_from_resi(sequence, seqx,seqy,resi_vec1,resi_vec2,byresi_opt); /* declare variable specific to this pair of TMalign */ double t0[3], u0[3][3]; double TM1, TM2; double TM3, TM4, TM5; // for a_opt, u_opt, d_opt double d0_0, TM_0; double d0A, d0B, d0u, d0a; double d0_out=5.0; string seqM, seqxA, seqyA;// for output alignment double rmsd0 = 0.0; int L_ali; // Aligned length in standard_TMscore double Liden=0; double TM_ali, rmsd_ali; // TMscore and rmsd in standard_TMscore int n_ali=0; int n_ali8=0; bool force_fast_opt=(getmin(xlen,ylen)>1500)?true:fast_opt; vector >tu_vec; vector do_vec; /* entry function for structure alignment */ int hingeNum=flexalign_main( xa, ya, seqx, seqy, secx, secy, t0, u0, tu_vec, TM1, TM2, TM3, TM4, TM5, d0_0, TM_0, d0A, d0B, d0u, d0a, d0_out, seqM, seqxA, seqyA, do_vec, rmsd0, L_ali, Liden, TM_ali, rmsd_ali, n_ali, n_ali8, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, force_fast_opt, mol_vec1[chain_i]+mol_vec2[chain_j],hinge_opt); if (hinge_opt && hingeNum<=1 && n_ali8<0.6*getmin(xlen,ylen)) { double t0_h[3], u0_h[3][3]; double TM1_h, TM2_h; double TM3_h, TM4_h, TM5_h; double d0_0_h, TM_0_h; double d0_out_h=5.0; string seqM_h, seqxA_h, seqyA_h; double rmsd0_h = 0.0; int L_ali_h; double Liden_h=0; double TM_ali_h, rmsd_ali_h; int n_ali_h=0; int n_ali8_h=0; vector >tu_vec_h(1,tu_vec[0]); vector do_vec_h; tu2t_u(tu_vec[0],t0_h,u0_h); int hingeNum_h=flexalign_main( xa, ya, seqx, seqy, secx, secy, t0_h, u0_h, tu_vec_h, TM1_h, TM2_h, TM3_h, TM4_h, TM5_h, d0_0_h, TM_0_h, d0A, d0B, d0u, d0a, d0_out_h, seqM_h, seqxA_h, seqyA_h, do_vec_h, rmsd0_h, L_ali_h, Liden_h, TM_ali_h, rmsd_ali_h, n_ali_h, n_ali8_h, xlen, ylen, sequence, Lnorm_ass, d0_scale, i_opt, a_opt, u_opt, d_opt, force_fast_opt, mol_vec1[chain_i]+mol_vec2[chain_j],hinge_opt); double TM =(TM1 >TM2 )?TM1 :TM2; double TM_h=(TM1_h>TM2_h)?TM1_h:TM2_h; if (TM_h>TM) { hingeNum=hingeNum_h; tu2t_u(tu_vec_h[0],t0,u0); TM1=TM1_h; TM2=TM2_h; TM3=TM3_h; TM4=TM4_h; TM5=TM5_h; d0_0=d0_0_h; TM_0=TM_0_h; d0_out=d0_out_h; seqM=seqM_h; seqxA=seqxA_h; seqyA=seqyA_h; rmsd0=rmsd0_h; L_ali=L_ali_h; Liden=Liden_h; TM_ali=TM_ali_h; rmsd_ali=rmsd_ali_h; n_ali=n_ali_h; n_ali8=n_ali8_h; tu_vec.clear(); for (int hinge=0;hinge1) { yname.clear(); for (chain_j=0;chain_j sequence; // get value from alignment file double Lnorm_ass, d0_scale; bool h_opt = false; // print full help message bool v_opt = false; // print version bool m_opt = false; // flag for -m, output rotation matrix int i_opt = 0; // 1 for -i, 3 for -I int o_opt = 0; // 1 for -o, 2 for -rasmol int a_opt = 0; // flag for -a, do not normalized by average length bool u_opt = false; // flag for -u, normalized by user specified length bool d_opt = false; // flag for -d, user specified d0 bool do_opt= false; // flag for -do, output distance of i-th aligned pair bool full_opt = false;// do not show chain level alignment double TMcut =-1; bool se_opt =false; int infmt1_opt=-1; // PDB or PDBx/mmCIF format for chain_1 int infmt2_opt=-1; // PDB or PDBx/mmCIF format for chain_2 int ter_opt =2; // END, or different chainID int split_opt =2; // split each chains int outfmt_opt=0; // set -outfmt to full output bool fast_opt =false; // flags for -fast, fTM-align algorithm int cp_opt =0; // do not check circular permutation int closeK_opt=-1; // number of atoms for SOI initial alignment. // 5 and 0 for -mm 5 and 6 int hinge_opt =9; // maximum number of hinge allowed for flexible int mirror_opt=0; // do not align mirror int het_opt=0; // do not read HETATM residues int mm_opt=0; // do not perform MM-align string atom_opt ="auto";// use C alpha atom for protein and C3' for RNA string mol_opt ="auto";// auto-detect the molecule type as protein/RNA string suffix_opt=""; // set -suffix to empty string dir_opt =""; // set -dir to empty string dirpair_opt=""; // set -dirpair to empty string dir1_opt =""; // set -dir1 to empty string dir2_opt =""; // set -dir2 to empty string chainmapfile=""; // chain mapping between two complexes int byresi_opt=0; // set -byresi to 0 vector chain1_list; // only when -dir1 is set vector chain2_list; // only when -dir2 is set vector chain2parse1; vector chain2parse2; vector model2parse1; vector model2parse2; vector > chain_pair_list; // only when -dirpair is set for(int i = 1; i < argc; i++) { if ( !strcmp(argv[i],"-o") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -o"); if (o_opt==2) cerr<<"Warning! -rasmol is already set. Ignore -o"<=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -rasmol"); if (o_opt==1) cerr<<"Warning! -o is already set. Ignore -rasmol"<=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -u or -L"); Lnorm_ass = atof(argv[i + 1]); u_opt = true; i++; if (Lnorm_ass<=0) PrintErrorAndQuit( "ERROR! The value for -u or -L should be >0"); } else if ( !strcmp(argv[i],"-a") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -a"); if (!strcmp(argv[i + 1], "T")) a_opt=true; else if (!strcmp(argv[i + 1], "F")) a_opt=false; else { a_opt=atoi(argv[i + 1]); if (a_opt!=-2 && a_opt!=-1 && a_opt!=1) PrintErrorAndQuit("-a must be -2, -1, 1, T or F"); } i++; } else if ( !strcmp(argv[i],"-full") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -full"); if (!strcmp(argv[i + 1], "T")) full_opt=true; else if (!strcmp(argv[i + 1], "F")) full_opt=false; else PrintErrorAndQuit("-full must be T or F"); i++; } else if ( !strcmp(argv[i],"-d") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -d"); d0_scale = atof(argv[i + 1]); d_opt = true; i++; } else if ( !strcmp(argv[i],"-closeK") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -closeK"); closeK_opt = atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-hinge") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -hinge"); hinge_opt = atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-v") ) { v_opt = true; } else if ( !strcmp(argv[i],"-do") ) { do_opt = true; } else if ( !strcmp(argv[i],"-h") ) { h_opt = true; } else if ( !strcmp(argv[i],"-i") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -i"); if (i_opt==3) PrintErrorAndQuit("ERROR! -i and -I cannot be used together"); fname_lign = argv[i + 1]; i_opt = 1; i++; } else if (!strcmp(argv[i], "-I") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -I"); if (i_opt==1) PrintErrorAndQuit("ERROR! -I and -i cannot be used together"); fname_lign = argv[i + 1]; i_opt = 3; i++; } else if (!strcmp(argv[i], "-chainmap") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -chainmap"); chainmapfile = argv[i + 1]; i++; } else if (!strcmp(argv[i], "-chain1") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -chain1"); split(argv[i+1],chain2parse1,','); i++; } else if (!strcmp(argv[i], "-chain2") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -chain2"); split(argv[i+1],chain2parse2,','); i++; } else if (!strcmp(argv[i], "-model1") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -model1"); split(argv[i+1],model2parse1,','); i++; } else if (!strcmp(argv[i], "-model2") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -model2"); split(argv[i+1],model2parse2,','); i++; } else if (!strcmp(argv[i], "-m") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -m"); fname_matrix = argv[i + 1]; m_opt = true; i++; }// get filename for rotation matrix else if (!strcmp(argv[i], "-fast")) { fast_opt = true; } else if (!strcmp(argv[i], "-se")) { se_opt = true; } else if ( !strcmp(argv[i],"-infmt1") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -infmt1"); infmt1_opt=atoi(argv[i + 1]); i++; if (infmt1_opt<-1 || infmt1_opt>3) PrintErrorAndQuit("ERROR! -infmt1 can only be -1, 0, 1, 2, or 3"); } else if ( !strcmp(argv[i],"-infmt2") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -infmt2"); infmt2_opt=atoi(argv[i + 1]); i++; if (infmt2_opt<-1 || infmt2_opt>3) PrintErrorAndQuit("ERROR! -infmt2 can only be -1, 0, 1, 2, or 3"); } else if ( !strcmp(argv[i],"-ter") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -ter"); ter_opt=atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-split") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -split"); split_opt=atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-atom") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -atom"); atom_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-mol") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -mol"); mol_opt=argv[i + 1]; i++; if (mol_opt=="prot") mol_opt="protein"; else if (mol_opt=="DNA") mol_opt="RNA"; if (mol_opt!="auto" && mol_opt!="protein" && mol_opt!="RNA") PrintErrorAndQuit("ERROR! Molecule type must be one of the " "following:\nauto, prot (the same as 'protein'), and " "RNA (the same as 'DNA')."); } else if ( !strcmp(argv[i],"-dir") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -dir"); dir_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-dirpair") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -dirpair"); dirpair_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-dir1") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -dir1"); dir1_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-dir2") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -dir2"); dir2_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-suffix") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -suffix"); suffix_opt=argv[i + 1]; i++; } else if ( !strcmp(argv[i],"-outfmt") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -outfmt"); outfmt_opt=atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-TMcut") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -TMcut"); TMcut=atof(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-byresi") || !strcmp(argv[i],"-tmscore") || !strcmp(argv[i],"-TMscore")) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -byresi"); byresi_opt=atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-seq") ) { byresi_opt=5; } else if ( !strcmp(argv[i],"-cp") ) { mm_opt=3; } else if ( !strcmp(argv[i],"-mirror") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -mirror"); mirror_opt=atoi(argv[i + 1]); i++; } else if ( !strcmp(argv[i],"-het") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -het"); het_opt=atoi(argv[i + 1]); i++; if (het_opt!=0 && het_opt!=1 && het_opt!=2) PrintErrorAndQuit("-het must be 0, 1, or 2"); } else if ( !strcmp(argv[i],"-mm") ) { if (i>=(argc-1)) PrintErrorAndQuit("ERROR! Missing value for -mm"); mm_opt=atoi(argv[i + 1]); i++; } else if (xname.size() == 0) xname=argv[i]; else if (yname.size() == 0) yname=argv[i]; else PrintErrorAndQuit(string("ERROR! Undefined option ")+argv[i]); } if (xname.size()==0 || (yname.size() && dir_opt.size()) || (yname.size() && dirpair_opt.size()) || (yname.size()==0 && dir_opt.size()==0 && dirpair_opt.size()==0)) { if (h_opt) print_help(h_opt); if (v_opt) { print_version(); exit(EXIT_FAILURE); } if (xname.size()==0) PrintErrorAndQuit("Please provide input structures"); else if (yname.size()==0 && dir_opt.size()==0 && dirpair_opt.size()==0 && mm_opt!=4) PrintErrorAndQuit("Please provide structure B"); else if (yname.size() && dir_opt.size()+dirpair_opt.size()) PrintErrorAndQuit("Please provide only one file name if -dir is set"); } if (suffix_opt.size() && dir_opt.size()+dirpair_opt.size()+dir1_opt.size()+dir2_opt.size()==0) PrintErrorAndQuit("-suffix is only valid if -dir, -dir1 or -dir2 is set"); if ((dir_opt.size() || dirpair_opt.size() || dir1_opt.size() || dir2_opt.size())) { if (mm_opt!=2 && mm_opt!=4) { if (o_opt) PrintErrorAndQuit("-o cannot be set with -dir, -dir1 or -dir2"); if (m_opt && fname_matrix!="-") PrintErrorAndQuit("-m can only be - or unset when using -dir, -dir1 or -dir2"); } else if ((dir_opt.size() || dirpair_opt.size() )&& (dir1_opt.size() || dir2_opt.size())) PrintErrorAndQuit("-dir cannot be set with -dir1 or -dir2"); else if (dir_opt.size() && dirpair_opt.size()) PrintErrorAndQuit("-dir cannot be set with -dirpair"); } if (o_opt && (infmt1_opt!=-1 && infmt1_opt!=0 && infmt1_opt!=3)) PrintErrorAndQuit("-o can only be used with -infmt1 -1, 0 or 3"); bool autojustify=(atom_opt=="auto" || atom_opt=="PC4'"); // auto re-pad atom name if (mol_opt=="protein" && atom_opt=="auto") atom_opt=" CA "; else if (mol_opt=="RNA" && atom_opt=="auto") atom_opt=" C3'"; if (atom_opt.size()!=4) { cerr<<"ERROR! Atom name must have 4 characters, including space.\n" "For example, C alpha, C3' and P atoms should be specified by\n" "-atom \" CA \", -atom \" P \" and -atom \" C3'\", respectively."<=5 || atom_opt.size()==0) return 1; else if (atom_opt.size()==1) atom_opt=" "+atom_opt+" "; else if (atom_opt.size()==2) atom_opt=" "+atom_opt+" "; else if (atom_opt.size()==3) atom_opt=" "+atom_opt; cerr<<"Change -atom to \""<0"); if (outfmt_opt>=2 && (a_opt || u_opt || d_opt)) PrintErrorAndQuit("-outfmt 2 cannot be used with -a, -u, -L, -d"); if (byresi_opt!=0) { if (i_opt) PrintErrorAndQuit("-byresi >=1 cannot be used with -i or -I"); if (byresi_opt<0 || byresi_opt>7) PrintErrorAndQuit("-byresi can only be 0 to 7"); if ((byresi_opt==2 || byresi_opt==3 || byresi_opt==6) && ter_opt>=2) PrintErrorAndQuit("-byresi 2 and 6 must be used with -ter <=1"); } //if (split_opt==1 && ter_opt!=0) //PrintErrorAndQuit("-split 1 should be used with -ter 0"); //else if (split_opt==2 && ter_opt!=0 && ter_opt!=1) //PrintErrorAndQuit("-split 2 should be used with -ter 0 or 1"); if (split_opt<0 || split_opt>2) PrintErrorAndQuit("-split can only be 0, 1 or 2"); if (mm_opt==3) { cp_opt=true; mm_opt=0; } if (cp_opt && i_opt) PrintErrorAndQuit("-mm 3 cannot be used with -i or -I"); if (mirror_opt && het_opt!=1) cerr<<"WARNING! -mirror was not used with -het 1. " <<"D amino acids may not be correctly aligned."<=2 && (mm_opt==1 || mm_opt==2)) PrintErrorAndQuit("-mm 1 or 2 must be used with -ter 0 or -ter 1"); if (mm_opt==4 && (yname.size() || dir2_opt.size())) cerr<<"WARNING! structure_2 is ignored for -mm 4"<=10) PrintErrorAndQuit("ERROR! -hinge must be <10"); if (chainmapfile.size() && mm_opt!=1) PrintErrorAndQuit("ERROR! -chainmap must be used with -mm 1"); /* read initial alignment file from 'align.txt' */ if (i_opt) read_user_alignment(sequence, fname_lign, i_opt); if (byresi_opt==6 || byresi_opt==7) mm_opt=1; else if (byresi_opt) i_opt=3; if (m_opt && fname_matrix == "") // Output rotation matrix: matrix.txt PrintErrorAndQuit("ERROR! Please provide a file name for option -m!"); /* parse file list */ int i; if (dirpair_opt.size()) file2chainpairlist(chain1_list,chain2_list, xname, dirpair_opt, suffix_opt); else { if (dir1_opt.size()+dir_opt.size()==0) chain1_list.push_back(xname); else file2chainlist(chain1_list, xname, dir_opt+dir1_opt, suffix_opt); if (dir_opt.size()) for (i=0;i tmp_vec1; vector tmp_vec2; for (i=0;i().swap(chain1_list); vector().swap(chain2_list); vector().swap(chain2parse1); vector().swap(chain2parse2); vector().swap(model2parse1); vector().swap(model2parse2); vector().swap(sequence); vector >().swap(chain_pair_list); t2 = clock(); float diff = ((float)t2 - (float)t1)/CLOCKS_PER_SEC; if (outfmt_opt<2) printf("#Total CPU time is %5.2f seconds\n", diff); return 0; }