Re: [AMBER] tleap box neutralization

From: Jason Swails <>
Date: Mon, 02 Mar 2015 08:55:57 -0500

The first time I tried sending this email, the uncompressed patch caused
the message to bounce. I'm sending it again (so the bounced email
waiting for approval can be discarded).

On Mon, 2015-03-02 at 09:13 +0000, Sergio Decherchi wrote:
> Dear Amber Users,
> I found that teLeap is not able to correctly neutralize the simulation box
> by counterions if the charge is in absolute value less than the nearest integer. That is
> 1.001 is correctly neutralized and 0.998 is not correctly neutralized.

This is actually only true for positive charges (i.e., when anions need
to be added for neutralization). By contrast, both charges -1.001 and
-0.998 result in 1 monovalent cation being added.

> As I see from the mailing list this is a well known issue.
> I also found that using lipids14 parameters (PA PC OL) this problem can arise (protein+memb
> system).
> Which is the source file (teLeap) that decides the number of counterions based
> on the total charge?

The file is commands.c. I actually adjusted this behavior back in
September of last year to always add the number of ions that will get
the charge closest to neutral. However, this fix is only present in the
developmental version of Amber (which will be released in April as the
next AmberTools release). In the meantime, I've attached a patch file
that implements these changes in AmberTools 14.

You can apply it by downloading it and using update_amber to apply it:

./update_amber --apply /path/to/leap.patch.bz2


Jason M. Swails
Rutgers University
Postdoctoral Researcher
Jason M. Swails
Rutgers University
Postdoctoral Researcher

AMBER mailing list

Received on Mon Mar 02 2015 - 06:00:04 PST
Custom Search