Re: [AMBER] Query regarding Linmin error

From: Jason Swails <>
Date: Thu, 20 Jan 2011 08:50:15 -0500

On Thu, Jan 20, 2011 at 12:58 AM, Hirdesh Kumar <>wrote:

> Hi Amber Users,
> I am doing MD of my protein bound with ligand and cofactor. The complex was
> prepared well using antechamber and tleap modules (did not show any
> error).*TIP3P
> * water type was used to solvate the system. I minimized the protein
> complex
> (*total :5320 atoms*) using steepest descent and conjugate gradient method
> (
> *Total: 2500 and ncyc:1500*). It showed "Linmin Error" when it moved from
> steepest descent to conjugate gradient. I searched for the error and found
> it to be related to some coordinate overlap. To solve the problem, I

This does not seem right to me (although maybe that can cause it to
happen...) LINMIN failures simply imply that the minimizer got stuck and
couldn't figure out how to find its way "downhill". If you've hit this
point in the minimization, then you are typically fine to start MD if that's
what you're aiming for. If you're really trying to find a *good* minimum,
then consider using the xmin minimizer (see the Amber manual, specifically
setting ntmin=3). Here's a page with more information .

> optimize my complex using Schrodinger and again prepare the protein complex
> using tleap. This time it is not showing such error (no warning in the
> mdout
> file) but still the job was killed at 1500 step. In mdout file also it is
> not showing any warning.

Was anything written to stderr?

Good luck!

> I am confused whether I should ignore this and optimize my system using
> increased number of steepest descent minimization and no conjugate gradient
> steps? Suggest me what should I do?
> Thanks,
> Hirdesh
> _______________________________________________
> AMBER mailing list

Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Graduate Student
AMBER mailing list
Received on Thu Jan 20 2011 - 06:00:04 PST
Custom Search