Re: [AMBER] Turning initial velocity off for continuous MD runs

From: David A Case <>
Date: Tue, 9 Mar 2021 16:12:03 -0500

On Tue, Mar 09, 2021, Aanshi Gandhi wrote:
>As a result, it would be meaningful for me to turn the ig flag
>“off” so that the initial velocities of each run is not determined at
>random but rather using the velocity from the former step. As per my
>understanding of reading the Amber manual, the two flags in addition to
>ig=0 that would affect this would be ntx=5 and irest=1.

Indeed, if you set ntx=5 and irest=1, the velocities will be taken from the
restart file.

>However even
>with these changes, it seems as though the initial velocities are chosen
>at random (based on how the RMSD plot looks).

Try a test where you set ntpr=1 and nstlim=10 and see what is happening.
Perhaps something else is wrong with your rather complex script. Run some
very short "chained" production runs by hand. Two short runs should give
you the same answer as a single longer run.

Since you not using ntt=3, the value of ig that you set should be ignored.


AMBER mailing list
Received on Tue Mar 09 2021 - 13:30:02 PST
Custom Search