Hi Baptiste
It's tough to say if there could be general parameters for computing
MMPBSA/MMGBSA. However, people have tried to develop guidelines to select
parameters for such calculations. For example,
https://pubmed.ncbi.nlm.nih.gov/21117705-assessing-the-performance-of-the-mmpbsa-and-mmgbsa-methods-1-the-accuracy-of-binding-free-energy-calculations-based-on-molecular-dynamics-simulations/
this
paper, to the best of my knowledge, is the first of its kind which attempts
to understand the effect of various simulation conditions (length of the
simulation) and various tunable parameters for MMGBSA/MMPBSA. I suggest you
thoroughly understand this paper.
I hope this helps.
Best Regards
On Thu, 6 Feb 2020 at 16:47, Baptiste Legrand <bap.legrand.gmail.com> wrote:
> Hi all,
>
> I just finished the tutorial about the MM/PB(GB)SA method with my
> complex. All is fine and using cuda, after some optimizations, the
> calculations could be launched on a workstation. I have just few general
> questions, reading the literature, one can found for similar complexes
> very different lengths of production run, from few ns to 1000 ns. I
> imagine that the error decrease with the length of the run and the
> number of frames but are there general/reasonable parameters? or each
> user should test different length of MD production before incorporating
> mutations etc.?
>
> Also, I found the tutorial to decompose the free energy contributions of
> each residue (quite redundant with the ala-scan?), but it seems to be
> appropriate only for amber11?
>
> https://ambermd.org/tutorials/advanced/tutorial3/py_script/section6.htm
>
> Thanks,
>
> Best regards,
>
> Baptiste
>
>
>
>
> _______________________________________________
> 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 Feb 06 2020 - 21:00:01 PST
