AMBER Archive (2004)

Subject: Re: AMBER: Do I need to repeat MD simulation?

From: Bill Ross (ross_at_cgl.ucsf.edu)
Date: Wed Jan 21 2004 - 10:01:54 CST

> (amber is single precision right?).

double precision

> Differences due to initial assignment of velocities

Another strategy could be to start from several different
points along a trajectory, reassigning velocities by different
seeds for a "fresher" start.

I think I remember a study from about 10 years ago showing
that multiple trajectories were better than longer simulations
at giving some confidence in free energy results.

Bill Ross

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu