You cannot speed it up; it is not parallelized. Antechamber is intended for monomers, not the polymer. The fact that it worked in the first place is due to increases in computing speeds of modern processors, but think of Antechamber, etc as products of the 90's. The intent is to parameterize the monomer (perhaps with blocking end groups that may have rules to fix the overall charge, etc.) and then build up the string of monomers rather than parameterizing the whole molecule. For proteins, the monomer is an amino acid, for example, blocked with Acetyl and N-Methyl on each end.
While there have been (not very successful or transferable) force fields parameterized based on the whole (folded) molecule, this is often not necessarily transferable to the unfolded state. In other words, as you parameterize to increasing polymer size, you may be introducing biases based on the initial structure that you parameterized to.
The GAFF / AMBER force field philosophy is to do multi-conformation fitting (and multi-orientation) on monomers -- noting that GAFF with RESP performs "better" than AM1-BCC for charges.
The way I would parameterize is to do an initial fit on the monomer (blocked ends, perhaps with constraints to make the blocked ends add up to zero charge with the blocked ends mimicking to some degree the chemistry / connectivity). Run MD on the monomer (GB or better in solution or in the representative surrounding environment), cluster, than re-fit the representative conformers in a multi-molecule RESP calculation. As RESP charges depend on a grid, do it with multi-orientations. The RED server, if alive, can help with the RESP part.
I cannot find a tutorial that illustrates this well; perhaps I should get back into the tutorials to help...
--tec3
________________________________________
From: Alesadi, Amirhadi <amirhadi.alesadi.ndsu.edu>
Sent: Thursday, April 22, 2021 9:26:23 PM
To: amber.ambermd.org
Subject: [AMBER] parmchk2
Hello,
I am Amirhadi Alesadi, PhD student of Civil Engineering at North Dakota State University, Fargo, ND.
I have a question about Ambertools.
I am using antechamer to generate the force field parameters of the polymer chain. For the chains composed of 3000 atoms or less, I successfully generated the "prmtop" and "inpcrd" files.
However, for larger system the following command takes much longer time:
parmchk2 -i my.mol2 -f mol2 -o my.frcmod
For running the "parmchk2" every time it takes more than 10 hours to finish the job.
This is my question, is there any way to increase the number of processors used for this job to speed up the code?
Regards,
Amirhadi
---
Amirhadi Alesadi
Ph.D. Student
Department of Civil & Environmental Engineering
North Dakota State University
NDSU Dept. 2470
PO Box 6050
Fargo, ND 58108-6050
amirhadi.alesadi.ndsu.edu
[cid:43f6df44-4020-4f69-80f0-8f199fdb1cee]
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Apr 22 2021 - 21:00:01 PDT
