I don't know the specifics of your situation, but as a GB model developer I
would definitely vote that PB is usually more reliable for this
application. We tend to make a lot of approximation sin GB so it can be
fast during MD, but for postprocessing without the need for forces I think
PB is the preferred approach.
Keep in mind that both models are sensitive to the input parameters (such
as the atomic radii set).
On Tue, Oct 20, 2020 at 8:28 AM <h_sahakyan.mb.sci.am> wrote:
> Dear Ray,
> Thank you very much for your help.
>
> As I know the default surface tension coefficient is different for GB and
> PB, I used default parameters for GB and inp=1 parameters for PB.
> Also, I performed energy decomposition and noticed that some amino acids,
> for instance, the N-terminal methionine, of GLY, PRO in some specific
> positions or the GTP molecule, can have very different contributions (~50
> fold), for more details see the attached files.
> Maybe this information can help solve this issue.
> Nevertheless, which method is more reliable, or how it is possible to
> validate this data? I used normal mode, but it is too slow since my protein
> complexes have about 13000 atoms, what would recommend for protein-protein
> interaction free energy calculation?
> Best regards,
> Harutyun
> _______________________________________________
> 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 Oct 20 2020 - 06:00:03 PDT
