Home › Forums › Zhang Lab Service System Discussion Board - New forum at https://zhanglab.dcmb.med.umich.edu/forum/

Modeling a 2200 residue protein

Sun, 06/21/2015 - 05:29 — arjunray

Hi,

I have a huge membrane protein (~2200 residues) that I am trying to model. In this regard, I had divided the protein according to the pfam domains (resulting in 3 segments) and successfully built the segments. Now I want to finally patch these three segments and build the whole structure but am limited by the sequence length range. I was wondering if there was any hack through which I could increase the limits?

Regards

Arjun

‹ I received notification email of job completion, but it is no result LOMETS2 hhsearch database error ›
Wed, 04/28/2021 - 04:57 — wonwill

a solution

This is a copy of the original post that was made by ITASSERteam at Mon, 06/22/2015 - 12:53. If this can be a problem, I will delete it.

----------

This is a typical problem of modeling multiple-domain proteins. I recommend the following procedure:

1. split sequence into domains and submit the sequence of individual domains to I-TASSER, like what you have done.
2. submit the whole-chain sequence to I-TASSER to get the whole-chain I-TASSER model. In case that the whole-chain sequence is beyond 1500, you can delete the N- and C-terminal sequence to make it below 1500 AA. But you should have the sequence include at least part of the sequence from each domain.
3. superpose the individual domain models to the whole-chain I-TASSER model to construct a new whole-chain structural model. You need to reorder the residue number of the single-domain models so that you can superpose the models appropriately to whole-chain model using TM-score program.
4. submit the new whole-chain model to FG-MD server to refine the model and remove steric clashes etc. Thus, you will a final full-length model that has all domain modeled appropriately.

This procedure can be quite complicated and tedious to beginning users. But this is probably the best way to model multi-domain proteins. We plan to make this procedure automated in the I-TASSER server in future.