Re: QM/MM MD and the qmParamPDB file

From: Marcelo C. R. Melo (melomcr_at_gmail.com)
Date: Wed Aug 15 2018 - 22:43:27 CDT

Dear Agisilaos

Internally, NAMD distributes atoms (both QM and MM) in "patches" that can
change order and exchange atoms, so the order in which atoms are processed
can change from step to step, without causing any problems. However, for
people doing advanced processing of QM results, this could be an issue, so
we changed this behavior to make sure that every QM input file (MOPAC, ORCA
or using a custom script) would be the same throughout the simulation, and
across simulations. This is implemented in the nightly build of NAMD. Is it
possible you are using the version 2.12 of NAMD, or earlier?

Best,
Marcelo

---
Marcelo Cardoso dos Reis Melo
PhD Candidate
Luthey-Schulten Group
University of Illinois at Urbana-Champaign
crdsdsr2_at_illinois.edu
+1 (217) 244-5983
On Wed, 15 Aug 2018 at 07:50, Agisilaos Chantzis <agisilaos.chantzis_at_gtn.ai>
wrote:
> Dear all,
>
>
> While running QM/MM+MD trajectories with NAMD and ORCA I have noticed
> that for each force calculation cycle the ordering of the QM atoms in
> the ORCA input file are changing. Although this seems not to alter the
> results (the molecule seems as it should be) I was wondering whether in
> a series of sequential QM/MM+MD jobs (e.g. minimization, annealing,
> equilibration etc) the PDB file specifying the QM atoms (the qmParamPDB
> file) should be the same in all configuration files or has to be changed
> to reflect the numbering of the QM atoms at the end of the previous run.
>
>
> I understand that for jobs without bonds being cut due to the QM/MM
> partition of the system this atoms reordering should not have any effect
> because the QM part will always remain the same, but what about the
> general case of cutting through bonds?
>
>
> Sincerely,
>
>
> Agisilaos
>
>

This archive was generated by hypermail 2.1.6 : Wed Sep 18 2019 - 23:20:14 CDT