Re: Computing interaction energies in namd

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Tue Feb 22 2011 - 19:07:11 CST

On Tue, Feb 22, 2011 at 7:21 PM, Samy Hamdouche <rich.fish_at_gmail.com> wrote:
> I've played around with the NAMDEnergy plugin of VMD. The problem with its
> application in my case is that it's very slow when computing on the order of
> 10000 interaction energies, because it creates 10000 separated pdb files
> with atom flags and runs a namd script with exactly the lines I quoted
> originally. I was wondering if there was a quicker (perhaps lower level) way
> of computing this many interaction energies within namd. Thanks,

why trying to make a 10 ton elephant like namd jump through hoops?

it would probably be easier to write a custom "pseudo MD" code that
just loops over the interaction tuples that you want. there are plenty
of "toy" MD codes around that can be adapted for that. at least that
is what i would suggest based on the very limited information about
your objective that you provide.

the next best thing would be to hack NAMD yourself, but i would
guess that it would take you longer.

axel.

> Samy
>
> On Tue, Feb 22, 2011 at 12:48 PM, JC Gumbart <gumbart_at_ks.uiuc.edu> wrote:
>>
>> Use the NAMDEnergy plugin in VMD.  You'll find it's very amenable to
>> scripting.
>>
>> On Feb 22, 2011, at 2:40 PM, Samy Hamdouche wrote:
>>
>> Hello,
>> I would like to compute the interaction energy between 4 atoms in a
>> protein, but I would like to do this for several different 4-atom clusters
>> in the protein. Within NAMD, the variable pairInteraction allows one to do
>> this given a PDB file with a flag for the desired four atoms. I.e.,
>> pairInteraction         on
>> pairInteractionGroup1 1
>> pairInteractionFile   namd-temp.pdb
>> pairInteractionSelf         on
>> However, for the number of four-atom groups for which I would like to
>> compute interaction energies (15,000), it would be unfeasible to create
>> 15,000 separate pdb files and read each one of them in the namd script. Is
>> there a simple way to compute interaction energies between given atoms in
>> namd where I can perhaps loop through all groups of atoms and print out the
>> within-group interaction energies for each group without creating 15,000
>> separate pdb files? I thought about just having 15000 different flags (1 for
>> each group), but this wouldn't work because many of the atoms are in
>> multiple groups. Thank you for your help,
>> Samy
>
>

-- 
Dr. Axel Kohlmeyer
akohlmey_at_gmail.com  http://goo.gl/1wk0
Institute for Computational Molecular Science
Temple University, Philadelphia PA, USA.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:23:37 CST