Re: [AMBER] PBSA: sasopt=0 incompatible with eneopt=4, frcopt=5

From: Ray Luo <rluo.uci.edu>
Date: Thu, 20 Oct 2016 09:22:40 -0700

Vlad,

Yes, it's a good idea to compare with different programs. Both our
group and many Amber users had been extensively explored different
programs before Amber10. In this regard, you probably want to try
Delphi first ... We have documented a test case showing how to
reproduce numerical results match those in the Delphi as much as
possible, the test case is $AMBERHOME/AmberTools/test/pbsa_delphi.
Apparently the numerical results are not exactly the same, but the
differences can be ignored when you use these setups for mmpbsa with
the two programs. Unfortunately this was for linear PB only. Delphi
does not support full nonlinear PB when we did the test. Not sure they
support it right now.

In general in these comparative studies, make sure you are using the
same input parameters, i.e. radii, charges, and molecular surface.
Also pay attention to the handling of interface jump conditions, i.e.
how solvent/solute dielectrics are handled on the molecular surface.
Finally convergence should be achieved with whatever you specified.

Grid spacing is probably less important in mmpbsa, but it doesn't hurt
to run your jobs with both 0.5 and 0.25 Angstrom. We showed grid
spacing does influences the numerical results in mmpbsa with PBSA, but
the numerical effect is very small with our default interface jump
condition treatment (controlled by smoothopt). Of course we only used
100 to 1000 snapshots in these comparisons, otherwise, the runs would
be too long. See DOI: 10.1002/jcc.24467. However, other smoothopts,
i.e. those that do not do much jump condition enforcement on the
molecular surface may not behave so good.

All the best,
Ray

--
Ray Luo, Ph.D.
Professor
Biochemistry, Molecular Biophysics, Chemical Physics,
Chemical and Biomedical Engineering
University of California, Irvine, CA 92697-3900
On Tue, Oct 18, 2016 at 2:23 PM, Vlad Cojocaru
<vlad.cojocaru.mpi-muenster.mpg.de> wrote:
> Thanks ! Maybe it would be good to add a note in the manual about this
> and also about other not obvious incompatible combinations.
>
> Best
> Vlad
>
> On 10/18/2016 05:39 PM, Ray Luo wrote:
>> Vlad,
>>
>> Yes, only the surface generated with the analytical density function
>> is supported with the new P3M approach, due to the force partition
>> algorithm used.
>>
>> All the best,
>> Ray
>> --
>> Ray Luo, Ph.D.
>> Professor
>> Biochemistry, Molecular Biophysics, Chemical Physics,
>> Chemical and Biomedical Engineering
>> University of California, Irvine, CA 92697-3900
>>
>>
>> On Tue, Oct 18, 2016 at 3:23 AM, Vlad Cojocaru
>> <vlad.cojocaru.mpi-muenster.mpg.de> wrote:
>>> Dear Ray, Dear Amber (PBSA) users,
>>>
>>> I noticed that whenever I try the option sasopt=0 (solvent excluded
>>> surface as molecular surfaces in PB calculations) in combination with
>>> eneopt=4 and frcopt=5 for non-linear PB I get a segmentation fault. When
>>> changing to eneopt=1 and frcopt=0, the calculation runs fine.
>>>
>>> I went back at the Wang et al paper 2012 where the sasopt=0 option is
>>> described but I could not find any particular hint on why this should be
>>> incompatible with eneopt=4 and frcopt=5 ... But sure, maybe (probably) I
>>> am missing something ....
>>>
>>> Is there any explanation for this behavior ?
>>>
>>> Thanks for answering
>>>
>>> Best wishes
>>> Vlad
>>>
>>> --
>>> Dr. Vlad Cojocaru
>>> Computational Structural Biology Laboratory
>>> Department of Cell and Developmental Biology
>>> Max Planck Institute for Molecular Biomedicine
>>> Röntgenstrasse 20, 48149 Münster, Germany
>>> Tel: +49-251-70365-324; Fax: +49-251-70365-399
>>> Email: vlad.cojocaru[at]mpi-muenster.mpg.de
>>> http://www.mpi-muenster.mpg.de/43241/cojocaru
>>>
>>>
>>> _______________________________________________
>>> 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
>
> --
> Dr. Vlad Cojocaru
> Computational Structural Biology Laboratory
> Department of Cell and Developmental Biology
> Max Planck Institute for Molecular Biomedicine
> Röntgenstrasse 20, 48149 Münster, Germany
> Tel: +49-251-70365-324; Fax: +49-251-70365-399
> Email: vlad.cojocaru[at]mpi-muenster.mpg.de
> http://www.mpi-muenster.mpg.de/43241/cojocaru
>
>
> _______________________________________________
> 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 Thu Oct 20 2016 - 09:30:03 PDT