Re: RE: Regarding Topology and Parameterization

From: SM Bargeen A Turzo (s.turzo_at_vikes.csuohio.edu)
Date: Wed Apr 11 2018 - 14:16:12 CDT

JC: yes I did load the psf file.

Here are the charges of the side chain I got from a QM ESP(Pop=MK) calculation:

CB 0.01
HB1 0.07
HB2 0.07
SG -0.15

________________________________
From: JC Gumbart <gumbart_at_physics.gatech.edu>
Sent: Wednesday, April 11, 2018 2:12:32 PM
To: NAMD list; Vermaas, Joshua
Cc: SM Bargeen A Turzo
Subject: Re: namd-l: RE: Regarding Topology and Parameterization

Actually, we simulated a radical state of a tyrosine here: https://www.ncbi.nlm.nih.gov/pubmed/28145121 (it’s artificially maintained in experiments)

Regarding the coloring in VMD, it may just be a scale issue. You could do

set sel [atomselect top “resname CYR”]
$sel get {name charge}

to see what charges VMD read from the PSF (you did load the PSF, right?).

Best,
JC

On Apr 11, 2018, at 12:02 PM, Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov<mailto:Joshua.Vermaas_at_nrel.gov>> wrote:

Hi Bargeen,

What is the charge on the radical? You should be able to see this in the psf. White just means that the charge is in the middle of the range, which I admit does not agree with what I think a radical looks like. In the past, I've seen folks do some basic QM to get an idea of what the charge distribution looks like on their radical states. However, I don't think I've seen extended MD on them, since radicals tend to be rather short-lived in reality, and MD timescales aren't always appropriate.

-Josh

On 2018-04-10 17:56:34-06:00 owner-namd-l_at_ks.uiuc.edu<mailto:owner-namd-l_at_ks.uiuc.edu> wrote:

Hi NAMD Users,
Somewhat new to NAMD, and I have this query: I am doing and MD simulation and my starting structure contains a cysteine radical and a small molecule. I tried editing the topology file and added another topology called CYR with appropriate charges so that NAMD would recognize that topology and would not automatically add the hydrogen. I also changed the particular residue to CYR so that they would match and PSF would be generated. This worked and the MD simulation was successful, however when I color the cysteine radical based on its charges in VMD, the side chain is completely white and looks like a dummy atom. Same thing is happening with the small molecule, that is 3 of the 22 atoms looks white. So my initial guess is that maybe the atom types are wrong but the small molecule stream file was generated with CGenFF. So I am not sure what the problem is. I checked the penalty section of str file(that is generated from CGenFF) and it is zero, so I am guessing I am good that side.
Here's what I am thinking in terms of solution and I was wondering if anyone can give me your opinion on it.
1) For the cysteine radical: I can prototnate the cysteine to cysteine+, generate the psf file and then manually adjust the charges on the side chain so that it is order with the atom types and the charges.
2) For the small molecule: it is part of a lipid so should I change the atomtypes in accordance to the lipid?
-Bargeen

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:59 CST