Re: [AMBER] Virtual box with imin = 1 or imin = 5

From: David Case <>
Date: Mon, 16 Dec 2019 20:15:34 +0000

On Mon, Dec 16, 2019, Jerome Eberhardt wrote:

>I tried to use igb = 6 with imin = 5, it does not work neither and if I
>do the minimization with imin = 1 and igb = 6, it gives me a completely
>different result at the end, with a positive final energy.

I'm a bit lost here: you show results for imin=1/igb=0/ntb=0, and for
imin=5/igb=0/ntb=0. I would have expected these to be the same, but
there may well be some bug here. Are both the prmtop file and all the
rst7 files non-periodic? I ask because you see to have WAT residues in
your system....

You say above that imin=5/igb=6/ntb=? does not work, but it would be
very helpful to see what the error is there.

You are exploring a corner case here: trying to apply a vacuum energy
function to a protein system is pretty unusual these days, and this
capability may have stopped working at some point. But using igb=6 (and
not igb=0) is definitely the way to go. (Setting igb=6 avoids the
formation of a "virtual box" and all its problems.)

...good luck....dac

AMBER mailing list
Received on Mon Dec 16 2019 - 12:30:02 PST
Custom Search