Re: Re: Problems to parametrize molecules

From: Mayne, Christopher G (
Date: Fri Apr 24 2015 - 14:22:55 CDT

There's a better (and easier) way than manually manipulating the data. I did miss one little part of my script that is required if you don't have chain/segname information in your PDB. Here's the easiest workflow, which is a minor modification to my previous one:

1) For some reason the resname in the topology definition provided in the CGenFF program (formerly ParamChem) renames the residue to the filename, not the resname in the MOL2. Modify the topology resname to UNK (i.e. matches the MOL2).

2) create a PDB from the MOL2,
mol new b_list.mol2
set sel [atomselect top all]
$sel writepdb b_list.pdb
$sel delete

3) use PSFGEN to create the file pair
package require psfgen
psfcontext reset
topology top_all36_cgenff.rtf
topology b_list.str
segment L { first none; last none; pdb b_list.pdb }
coordpdb b_list.pdb L
regenerate angles dihedrals
writepsf b_list.psf

NOTE the addition of the "L" at the end of the "coordpdb b_list.pdb L" line. This matches the coordinates to the L segment, created above. If the L is missing, PSFGEN doesn't know where to use the coordinates. This was the minor mistake in the previous instructions.


On Apr 24, 2015, at 2:02 PM, Bennion, Brian wrote:<> use to give out the already updated mol2 in addition to the str (par/top) files. I am not sure why that functionality was removed.

This would be very helpful in the future (although the funding for paramchem has been altered/cut)

From: Mayne, Christopher G [<>]
Sent: Friday, April 24, 2015 11:53 AM
To: Evandro Semighini
Cc: Bennion, Brian;<>
Subject: Re: namd-l: Re: Problems to parametrize molecules

Can we also have the ParamChem output as well?

Christopher Mayne

On Apr 24, 2015, at 12:49 PM, Evandro Semighini wrote:

Hello Brian,

The molecule is attached to the e-mail, in mol2 format.

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:05 CST