[gmx-users] QM/MM simulations

[Justin Lemkul]

On 4/29/14, 4:46 AM, Iamkaant wrote:
> So, I'v made a little test run based on slightly modified  tutorial
> <http://wwwuser.gwdg.de/~ggroenh/EMBO2004/html/qmmmvacuum.html#top>  files.
> As far as I can judge, mdrun operates properly, ORCA is called and also
> works OK.
> Here is zipped working directory.  3.ZIP
> <http://gromacs.5086.x6.nabble.com/file/n5016133/3.ZIP>
> One thing that bothers me is sluggishness of ORCA optimization steps -- they
> took probably 99.9% of wall-time. Earlier I'v made optimizations by this
> instance of ORCA and they went very fast, so the issue may be in the input
> parameters. But since I'm not very familiar with it, I can't say what slows
> down calculation for system of less than 30 atoms at RHF/3-21G level.

IIRC, that is to be expected.  mdrun simply serves as a wrapper to pass 
information back and forth with ORCA, and the QM part of the calculation in 
Gromacs QM/MM can only be executed in serial, so it will be very slow.

Thanks for reporting back; it's good to know that things are working (even if 
performance could stand to be improved).

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================