Re: Restarting an alchemic FEP calculation

From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Mon Jun 27 2011 - 11:03:55 CDT

Hi Giovanni,

The simplest thing to do is to restart from the beginning of the
lambda window that was interrupted. You will find that information in
the standard output (messages when changing lambda values) or in the
fepout file (lines starting with '#'). Then use the standard NAMD
restart files (coor, vel, xsc). At least, this is correct if the
restart frequency is smaller than the sampling time for a lambda
window. If not, you may need to add more time to equilibrate the
system at the current lambda value at the beginning of the new
simulation.

Finally, you'll need to sum the free energy difference for all windows
from both files. Exactly how to do that depends a little on what
software you want to use.

See also:
http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/9511.html
http://www.ks.uiuc.edu/Research/namd/mailing_list/namd-l/6834.html

Cheers,
Jerome

On 27 June 2011 16:32, Giovanni Bellesia <giovanni.bellesia_at_gmail.com> wrote:
> Hi all,
>
> an alchemic FEP calculation I was running stopped a couple of
> lambda-steps from the end due to a power outage.
> Is there a safe way to restart it and complete it making use of the
> free energy data that I already have ?
>
> Thanks
>
> Giovanni
>
>

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:20:30 CST