- Posts: 90
martinize.py bug
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
I tried to martinize a protein complex with 7 chains in it (chains A-G). The last seven residues of chain G are missing in the .itp file, i.e., Protein_G.itp, whereas the rest (Protein_A.itp-Protein_F.itp) have all the residues. I checked my dssp file and my input coordinate file (which I have used for AA simulations as well). They seem to be fine. Can anyone help me out with this?
Thanks,
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
Please Log in or Create an account to join the conversation.
- djurre
- Offline
- Admin
- Posts: 272
I don't remember exactly but there is/was a problem with do_dssp and multiple chains. The string of SS characters is supposed to be splitted by equals signs (=), but this was in some versions frameshifted by one amino acid. Could that be the problem?
In that case switching to a newer gromacs (do_dssp) or calling dssp directly from martinize.py (e.g.: -dssp /usr/bin/dssp) should fix the problem.
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
Neither strategies worked. I am using gromacs 4.6.4, but I am not sure what do_dssp has got to do with martinize.py
I passed the dssp executable (various versions, new to old) directly to martinize.py, as well as passed the .dssp files directly. Nothing works. The last 7 residues of the 7th chain remain missing!
Interestingly, though the chains are supposed to be identical, the force-field parameters (equilibrium distances for bonds and angles) in the .itp files for the individual chains (say A.itp versus D.itp) are slightly different. Why is that? This didn't happen for another multi-chain protein that I ran Martini simulations with.
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
I am willing to share my files with someone if necessary. I think that this is may be a bug with martinize.py
Thank you.
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
It might have been the atom name of the termini in the pdb file that did not match the ones defined in the martinize.
Rajat Desikan wrote: Update: I tried with an earlier version of martinize.py (Aug 13) that had successfully worked with another protein. I still have the same problem of missing residues.
I am willing to share my files with someone if necessary. I think that this is may be a bug with martinize.py
Thank you.
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
The other chains have termini atoms too, right? The .itp files are just fine for those.
Is there a solution for this? Do you have a corrected version of martinize.py?
Thanks,
Please Log in or Create an account to join the conversation.
- djurre
- Offline
- Admin
- Posts: 272
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
It is a modified version of a PDB file with reconstructed residues. Can I mail you the .pdb and any other files that you may require?
Thanks,
Please Log in or Create an account to join the conversation.
- djurre
- Offline
- Admin
- Posts: 272
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
I love when things get fixed.
Could we get this on the list or the next martinize version?
Tks
Rajat Desikan wrote: Thank you Djurre, Xavier. That worked wonderfully!!
Please Log in or Create an account to join the conversation.
- djurre
- Offline
- Admin
- Posts: 272
Please Log in or Create an account to join the conversation.
- Rajat Desikan
- Topic Author
- Offline
- Expert Boarder
- Posts: 90
Please Log in or Create an account to join the conversation.
- xavier
- Offline
- Admin
- Posts: 416
Rajat Desikan wrote: I had processed the .pdb with pdb2gmx (amber99sb-ildn ff), so I assumed that the termini were properly capped!
Please Log in or Create an account to join the conversation.