SPICKER is a clustering algorithm to identify the near-native models from
a pool of protein structure decoys.
You can install and run the SPICKER program at your
own computers by
- 1, download and unpack
This package includes both source code
and executable program
as well as other illustration files.
- 2, Although you can run ./spicker directly, it is recommended that you recompile 'spicker.f'
on your computer, which should increase the speed of the program.
The suggested compiler include (depending on available compilers in your computer)
>gfortran -static -O3 -ffast-math -lm -o spicker spicker.f
>g77 -static -O3 -o -lm spicker spicker.f
For more detail,
please read the attached readme file
You can find a set of I-TASSER decoys
to test the SPICKER program.
spicker_v1.0 was released. In this version, a bug with regard to the input format
spicker_v1.0 package can be downloaded at
SPICKER V2.0 was released. In this version, a new option is added to
automatically tune RMSD cutoff based on variation of
decoy distributions. This option works best for decoys
generated by ab initio simulations, such as QUARK.
spicker_v2.0 package can be downloaded at
- 2010/09/14: modified 'seq.dat' to a simplied format
- 2010/12/29: modified 'rmsinp' to a simplied format
- 2014/01/09: add a license note to 'spicker.f' and modify the instruction part
SPICKER V3.0 was released. In this version,
an option was added to cluster decoys using 1/TM-score to replace
RMSD as the distance scale. This part is written by
Dr. Jens Thomas.
spicker_v3.0 package can be downloaded at
Y. Zhang, J. Skolnick, SPICKER: A Clustering Approach to Identify Near-Native Protein Folds
, Journal of Computational Chemistry, 2004 25: 865-871.
(download the PDF file