RE: What is the exact form of the harmonic potential used by the NAMD "constraints" option?

From: Lennart Nilsson (Lennart.Nilsson_at_ki.se)
Date: Tue Sep 19 2017 - 12:29:33 CDT

With an exponent that may be different from 2, the factor 0.5 becomes rather arbitrary, which is why we (CHARMM) dropped it.

Best wishes,
Lennart Nilsson

From: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] On Behalf Of Giacomo Fiorin
Sent: den 19 september 2017 19:24
To: NAMD list <namd-l_at_ks.uiuc.edu>; Brian Radak <bradak_at_anl.gov>
Cc: Victoria Lim <limvt_at_uci.edu>
Subject: Re: namd-l: What is the exact form of the harmonic potential used by the NAMD "constraints" option?

NAMD follows the CHARMM convention for this. Although I'm not familiar with the CHARMM internal code, they do report the formula without the 1/2 that Brian already pasted.

The Colvars manual specifies its harmonic formula explicitly to avoid confusion, particularly because of the vague NAMD definition of the keyword.

Giacomo

On Tue, Sep 19, 2017 at 12:39 PM, Brian Radak <bradak_at_anl.gov<mailto:bradak_at_anl.gov>> wrote:

Yes - but that is only in the memory optimized version, which is barely used except for extremely large systems on large supercomputers.

In the vast majority of instances, the force constant is just the value from the PDB file, usually stored as the B-factor. This is what is indicated by the manual.

On 09/19/2017 12:33 PM, Victoria Lim wrote:
I looked through a few versions of the user guide (such as http://www.ks.uiuc.edu/Research/namd/2.11/ug/node27.html) but did not see the particular formula used.
Thanks for looking through the code Brian! Out of curiosity, when you say the force constants are read as simple Boolean flags, do you mean that if k is zero, the Boolean variable goes to false, and the atom is unrestrained?


On Tue, Sep 19, 2017 at 6:30 AM, Brian Radak <bradak_at_anl.gov<mailto:bradak_at_anl.gov>> wrote:

This a worthwhile question - is this not in the user guide? I would regard that as an oversight.

Looking at the code, the energy is computed as:

E = k(r - r_ref)**n

where n is the (positive integer) exponent and defaults to 2.

Something I did not now, but is probably of little consequence, is that the memory optimized version of NAMD appears to ignore the force constants in the reference file and instead reads them as simple boolean flags. Otherwise k is computed as the value in the given PDB column times constraintScaling.

HTH,

Brian

On 09/19/2017 02:37 AM, Victoria Lim wrote:
Dear NAMD users,

Do the harmonic restraint parameters specified by the "constraints" option contain the factor of one half in the harmonic potential? I realize the "extraBonds" option does not contain this factor, whereas the harmonic potential in the colvars module does have the 1/2. However, I could not find an explicit formula for the constraints option.

Thank you,
Ria

--
Brian Radak
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory

9700 South Cass Avenue, Bldg. 240<https://maps.google.com/?q=9700+South+Cass+Avenue,+Bldg.+240&entry=gmail&source=g>
Argonne, IL 60439-4854
(630) 252-8643<tel:%28630%29%20252-8643>
brian.radak_at_anl.gov<mailto:brian.radak_at_anl.gov>


--
Brian Radak
Postdoctoral Appointee
Leadership Computing Facility
Argonne National Laboratory

9700 South Cass Avenue, Bldg. 240<https://maps.google.com/?q=9700+South+Cass+Avenue,+Bldg.+240&entry=gmail&source=g>
Argonne, IL 60439-4854
(630) 252-8643<tel:(630)%20252-8643>
brian.radak_at_anl.gov<mailto:brian.radak_at_anl.gov>



--
Giacomo Fiorin
Associate Professor of Research, Temple University, Philadelphia, PA
Contractor, National Institutes of Health, Bethesda, MD
http://goo.gl/Q3TBQU
https://github.com/giacomofiorin

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2018 - 23:20:36 CST