Re: [AMBER] NaN VDWAALS Error During Minimization

From: David Cerutti <dscerutti.gmail.com>
Date: Tue, 5 Jan 2021 20:29:19 -0500

It looks like two of your atoms are somehow overlapping, which may not be
obvious from a visual inspection of a complex system. One clue is the
maximum force you are seeing--it identifies an atom experiencing great
strain. You could check that. Alternatively, you could try loading your
system into mdgx and running a "configuration sampling" of a single
conformation. The input would look something like this:

&files
  -p YourTopology
  -c YourInpcrd
  -o LookAtThis.out
&end

&configs

  % General minimization instructions
  count = 1,
  verbose 1,
  maxcyc = 0,
  strainlim = 20.0,
  bstrain = 10.0,
  astrain = 10.0,

  % Output controls
  outbase 'Conf', 'Conf',
  write 'cdf', 'pdb',
  outsuff 'cdf', 'pdb',
&end

The point is to run one evaluation of the energy (maxcyc = 0 I think will
do it) and then let the routines for checking strain between any given pair
of atoms go to work and tell you where the strain lies. There will be some
detail printed in LookAtThis.out afterwards, although this will be an
all-to-all pair calculation and therefore may take some time to run. I
have used this strategy in the past to elucidate what region of my system
is to blame when minimization fails.

The more general purpose of these features is to minimize hundreds of
copies of a SMALL molecule at the same time, all subject to different
restraints to make their conformations diverge, and then analyze any
results that failed to reach a local energy minimum without great strain in
some bond, angle, or overall restraint energy. It will give somewhat more
detailed explanations of these failed systems than the imin=1 diagnostics
in a sander mdout, so that the user can then tailor the restraints or add
other minimization attempts to improve the success rate and ultimately get
a large set of energy-minimized conformations which are geometrically
distinct but all reasonably relaxed.

Dave


On Tue, Jan 5, 2021 at 7:55 PM Jatin Kashyap <jk435.njit.edu> wrote:

> Dear AMBER Community,
>
> I am getting an error[1] upon performing minimization of drug-protein
> complex by using the AMBER tutorial[2].
> There seems to be nothing wrong with the complex as per visual inspection
> of complex PDB file obtained after parametrization.
> Please let me know if sharing the output file can be more helpful.
>
> Thank you very much.
>
> [1]
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 1 NaN NaN 9.2217E+09 O1 18
>
> BOND = 13750.1241 ANGLE = 10976.5308 DIHED =
> 2934.9414
> VDWAALS = NaN EEL = Infinity EGB =
> 60228.2741
> 1-4 VDW = 381037.7111 1-4 EEL = 14860.5376 RESTRAINT =
> 0.0000
> .
> .
> .
> .
> .
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 2519 NaN NaN 0.0000E+00 N 1
>
> BOND = NaN ANGLE = 607581.2350 DIHED =
> 2262.8000
> VDWAALS = 0.0000 EEL = 0.0000 EGB =
> -2653355.7457
> 1-4 VDW = NaN 1-4 EEL = NaN RESTRAINT =
> 0.0000
>
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 2520 NaN NaN 0.0000E+00 N 1
>
> BOND = NaN ANGLE = 607581.2350 DIHED =
> 2262.8000
> VDWAALS = 0.0000 EEL = 0.0000 EGB =
> -2653355.7457
> 1-4 VDW = NaN 1-4 EEL = NaN RESTRAINT =
> 0.0000
>
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 2521 NaN NaN 0.0000E+00 N 1
>
> BOND = NaN ANGLE = 607581.2350 DIHED =
> 2262.8000
> VDWAALS = 0.0000 EEL = 0.0000 EGB =
> -2653355.7457
> 1-4 VDW = NaN 1-4 EEL = NaN RESTRAINT =
> 0.0000
>
> ... RESTARTED DUE TO LINMIN FAILURE ...
>
> ***** REPEATED LINMIN FAILURE *****
>
> ***** SEE http://ambermd.org/Questions/linmin.html FOR MORE INFO
> *****
>
>
> FINAL RESULTS
>
>
>
> NSTEP ENERGY RMS GMAX NAME NUMBER
> 2521 NaN NaN 0.0000E+00 N 1
>
> BOND = NaN ANGLE = 607581.2350 DIHED =
> 2262.8000
> VDWAALS = 0.0000 EEL = 0.0000 EGB =
> -2653355.7457
> 1-4 VDW = NaN 1-4 EEL = NaN RESTRAINT =
> 0.0000
>
>
> [2]
> https://ambermd.org/tutorials/basic/tutorial4b/index.php
>
> ——
> Jatin Kashyap
> Ph.D. Student
> Department of Mechanical and Industrial Engineering
> New Jersey Institute of Technology (NJIT)
> University Heights
> Newark, NJ 07102-1982
> Phone- (201)889-5783
> Email- jk435.njit.edu
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Jan 05 2021 - 17:30:02 PST
Custom Search