Re: [AMBER] Multiple Replicate Trajectories

From: David A Case <>
Date: Wed, 3 Nov 2021 11:52:16 -0400

On Wed, Nov 03, 2021, Adrian Roitberg wrote:
>irest=1,ntx=5 will pick up coordinates AND velocities from the initial
>structure. This means you start your MD at the right temperature, and
>then ig=1, ntt=3 will create a different random number sequence for
>each runs, which will make them "independent".
>irest=0, ntx=1 with ntt=3, ig=1 will also make them independent, but
>now the initial velocities will be random, and there is a small chance
>that the initial T would be different than your request target. This
>is not crucial, since it goes to the proper one reasonably fast.

Important: you want to set ig=-1 (*not* ig=1) to get the behavior that
Adrian is describing. (ig=-1 is the default, if you don't otherwise specify


AMBER mailing list
Received on Wed Nov 03 2021 - 09:00:02 PDT
Custom Search