In using mcfgen.py -f
- shige
-
Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
6 years 1 month ago #469
by shige
Dear Cassandra users,
I’m trying to generate some randomized conformations of a small molecule (tmcmpd_h1_1.pdb, as attached).
I used mcfgen.py with –f option to read Gromacs topology files (MOL01_GMX.top and MOL01_GMX.itp, as attached).
Then, libraly_setup.py seems to successfully generate the fragment libraries.
When I executed Cassandra using nvt.inp, the structures of the molecule were corrupted both at the beginning (step 0) and the end (step 10000) , where the –CHO fragment jumped far apart from the remaining moieties (nvt.out.xyz, as attached).
The top and itp files were created using AMBER acpype.py. Is this problem purely due to the force field parameters?
Let me attach the followings.
tmcmpd_h1_1.pdb
MOL01_GMX.top
MOL01_GMX.itp
nvt.inp
nvt.out.xyz
tmcmpd_h1_1-out.tiff (Figure of the corrupted structure)
tmcmpd_h1_1.mcf
Thanks for kindful suggestions,
--Yasuhiro
I’m trying to generate some randomized conformations of a small molecule (tmcmpd_h1_1.pdb, as attached).
I used mcfgen.py with –f option to read Gromacs topology files (MOL01_GMX.top and MOL01_GMX.itp, as attached).
Then, libraly_setup.py seems to successfully generate the fragment libraries.
When I executed Cassandra using nvt.inp, the structures of the molecule were corrupted both at the beginning (step 0) and the end (step 10000) , where the –CHO fragment jumped far apart from the remaining moieties (nvt.out.xyz, as attached).
The top and itp files were created using AMBER acpype.py. Is this problem purely due to the force field parameters?
Let me attach the followings.
tmcmpd_h1_1.pdb
MOL01_GMX.top
MOL01_GMX.itp
nvt.inp
nvt.out.xyz
tmcmpd_h1_1-out.tiff (Figure of the corrupted structure)
tmcmpd_h1_1.mcf
Thanks for kindful suggestions,
--Yasuhiro
Please Log in to join the conversation.
- jshah46545
-
- Offline
- Junior Member
-
Less
More
- Posts: 27
- Thank you received: 5
- shige
-
Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
- ryangmullen
-
- Offline
- Administrator
-
Less
More
- Posts: 124
- Karma: 4
- Thank you received: 24
6 years 1 month ago #472
by ryangmullen
Hi Yasuhiro,
It looks like there's a bug in mcfgen.py. Atom 2 is being misidentified as a ring atom instead of an exoring atom. As a result, only 7 fragments are being identified instead of 8. With the current scripts, you can circumvent this by putting your 6 ring atoms sequentially in the *pdb (and accompanying *itp). Or we'll notify you when we've patched mcfgen.py.
Also, please note the following (from mcfgen.py -h):
2) This script does not currently support multiple dihedrals for the same 4
atoms. If mulitple parameters are given for the same dihedral sequence,
only the last parameters will be used.
3) Improper definitions are not currently read from Gromacs forcefield files.
Your *itp defines two dihedral parameter sets for the same 4 atoms (eg 14 11 13 19). Only 1 of these is in your *mcf. None of the impropers from the last section of your *itp are in the *mcf.
Ryan
It looks like there's a bug in mcfgen.py. Atom 2 is being misidentified as a ring atom instead of an exoring atom. As a result, only 7 fragments are being identified instead of 8. With the current scripts, you can circumvent this by putting your 6 ring atoms sequentially in the *pdb (and accompanying *itp). Or we'll notify you when we've patched mcfgen.py.
Also, please note the following (from mcfgen.py -h):
2) This script does not currently support multiple dihedrals for the same 4
atoms. If mulitple parameters are given for the same dihedral sequence,
only the last parameters will be used.
3) Improper definitions are not currently read from Gromacs forcefield files.
Your *itp defines two dihedral parameter sets for the same 4 atoms (eg 14 11 13 19). Only 1 of these is in your *mcf. None of the impropers from the last section of your *itp are in the *mcf.
Ryan
The following user(s) said Thank You: shige
Please Log in to join the conversation.
- shige
-
Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
6 years 1 month ago #473
by shige
Thank Jindal & Ryan for your useful advices,
I tried two indexing cases (rev and rev5, as attached) for the molecule with the ring number rearranged in “sequential” (it means continuous?) order both for the *pdb and *itp. The generated structures, however, remained corrupted, where the molecules were fragmented into 7 pieces, not 8. I revised by hand # fragment_info and # fragment_connectivity in *mcf and tried again, but the job halted with segmentation fault.
Anyway, I’ll wait for the bugfix version of mcfgen.py.
I would be glad with further advices.
Yasuhiro
I tried two indexing cases (rev and rev5, as attached) for the molecule with the ring number rearranged in “sequential” (it means continuous?) order both for the *pdb and *itp. The generated structures, however, remained corrupted, where the molecules were fragmented into 7 pieces, not 8. I revised by hand # fragment_info and # fragment_connectivity in *mcf and tried again, but the job halted with segmentation fault.
Anyway, I’ll wait for the bugfix version of mcfgen.py.
I would be glad with further advices.
Yasuhiro
Please Log in to join the conversation.
- ryangmullen
-
- Offline
- Administrator
-
Less
More
- Posts: 124
- Karma: 4
- Thank you received: 24
6 years 1 month ago - 6 years 1 month ago #475
by ryangmullen
Hi Yasuhiro,
I'm not sure how quickly we'll be able to turn around the bug fix on mcfgen.py. In the meantime, I've attached input files for your system that run without error. Please let us know if you experience any further issues.
Here's a brief summary of the changes I made to your files:
Ryan
I'm not sure how quickly we'll be able to turn around the bug fix on mcfgen.py. In the meantime, I've attached input files for your system that run without error. Please let us know if you experience any further issues.
Here's a brief summary of the changes I made to your files:
- inp.inp - Cassandra input file. The key change I made was to switch the Vdw_Style and Charge_Style from "minimum_image" to "cut" with a 25 Angstrom (half-box) cutoff.
- inp.pdb - input PDB file. I changed the ordering of atoms on line 37 from "CONECT 1 7 2 3" to "CONECT 1 2 3 7"
- from_mcfgen.mcf - mcf generated using the command "mcfgen.py -f inp.top inp.pdb". Note that this mcf identifies 8 fragments, but it is still missing 2 dihedral parameter sets and all the impropers from the inp.itp
- inp.mcf - I modified from_mcfgen.mcf to add the missing dihedrals and impropers noted above.
Ryan
Last Edit: 6 years 1 month ago by ryangmullen.
The following user(s) said Thank You: shige
Please Log in to join the conversation.
Time to create page: 0.149 seconds