Re: Multiple Colvars

From: McGuire, Kelly (mcg05004_at_byui.edu)
Date: Wed Jul 25 2018 - 10:26:46 CDT

I was able to get a distanceZ + one dihedral colvar to work (so, 2D colvar), it appears that the atoms moving too fast was due to not selecting the atomnumbers in the correct order and/or incorrectly selecting one of the atom serial numbers.

The memory error appears to be what Joshua suggested, not closing of one of the brackets and getting, essentially, an infinite loop using up all of the memory.

Now, Giacomo or Joshua, since this is now a 2D (eventually going to 5D or more) PMF in that window, I cannot merge it with the other 1D windows, correct? Do I, therefore, need to add these dihedral colvars to all of my windows to make them the same dimension in order to merge them into one PMF? Thanks!

Kelly L. McGuire

PhD Scholar

Biophysics

Department of Physiology and Developmental Biology

Brigham Young University

LSB 3050

Provo, UT 84602

________________________________
From: Vermaas, Joshua <Joshua.Vermaas_at_nrel.gov>
Sent: Tuesday, July 24, 2018 3:59:23 PM
To: McGuire, Kelly; namd-l_at_ks.uiuc.edu
Subject: RE: Multiple Colvars

I've had several hundred colvars in a configuration before, so in principle what you are doing ought to work, with the caveat that your colvar file needs to be parseable. I vaguely recall seeing this once when I didn't close a component definition, and the parser ran itself out of memory trying to find the end of the configuration file.

-Josh

On 2018-07-24 14:43:55-06:00 owner-namd-l_at_ks.uiuc.edu wrote:

I'm still learning how to use colvars. As stated in other posts, the free energy in bulk water on one side of my membrane ion channel is about 30 kcal/mol higher than bulk water on the other side of the channel. We think this discrepancy is due to other orthogonal degrees of freedom needing to be biased. We would like to add a bias to some of the dihedrals in this channel. I have my usual distanceZ colvar set for the center of mass of the channel and the center of mass of the ligand.

Now, I am trying to add colvars for the dihedrals of four sidechains (channel is an homotetramer), the four histidine sidechains. That would be 5 colvars in my file. When I run the ABF, I get the error:

FATAL ERROR: Memory allocation failed on processor 0

If, I only do my distanceZ colvar and one of the dihedral colvars, then that error goes away, and then I get the error atom moving too fast...

What is causing both of those errors? Can I not have multiple dihedral colvars in the same file?

Kelly L. McGuire

PhD Scholar

Biophysics

Department of Physiology and Developmental Biology

Brigham Young University

LSB 3050

Provo, UT 84602

This archive was generated by hypermail 2.1.6 : Tue Dec 10 2019 - 23:20:01 CST