Hi Chris,
You're right that gb has lots of options, that's why I suggested checking
what people like Rob Rizzo use to get good match to expt.
I agree about the desire for negative binding free energy, just wanted to
point out that you should be open to other issues as well (initial
structure, sampling, ligand parameters, etc). People seem to get better
results for relative affinities using mm-gbsa rather than absolute.
Carlos
On Jun 18, 2015 8:30 AM, "Chris" <dlutlife223.163.com> wrote:
> Dear Carlos:
> Thx for reply. A negative dG should explain favorable interaction between
> ligand and protein, so, I think, PBSA gives a reasonable approximate
> energy, althought -TdS is not included.
> On the other hand, GBSA is more approximate, but, I think it SHOULD give a
> correct property of energy, a negative number. Because, positive and
> negative of energy represented
> a totally different property of system.
>
>
> What I am doubted is if I am using a improper parameters for GBSA
> calculation or just GBSA is not suitable for my system.
>
> Yours sincerely
> Chris
>
>
>
>
>
>
>
>
> At 2015-06-18 20:05:11, "Carlos Simmerling" <carlos.simmerling.gmail.com>
> wrote:
> >Chris,
> >If PB works well (although it's not clear to me how you have determined
> >that the PB result is "better"), why try to use GB? For what you are doing
> >I usually prefer PB. GB is more approximate but faster so it has the
> >advantage for use during md, but not for post-processing.
> >
> >Note also that there is a significant difference in these 2 calculations I
> >your nonpolar solvation result.
> >
> >You might look into the literature for what options people use with gb for
> >doing mm-gbsa validated against experiment, for example work by Rob Rizzo.
> > On Jun 18, 2015 5:33 AM, "Chris" <dlutlife223.163.com> wrote:
> >
> >> Deal All!
> >>
> >> I test my system with MM-GBSA and MM-PBSA both:
> >> ################################################
> >> &general
> >> startframe=3641,endframe=4000, verbose=1,
> >> /
> >> &gb
> >> igb=2, saltcon=0.100,
> >> /
> >> &pb
> >> istrng=0.15,radiopt=0,
> >> /
> >> &decomp
> >> idecomp=1, print_res="1-252"
> >> dec_verbose=0,
> >> /
> >> #############################################
> >> and I use "set default PBRadii mbondi2" to prepare my prmtop file.
> >> All goes well, and result belows:
> >>
> >> GBSA:
> >>
> >> Differences (Complex - Receptor - Ligand):
> >> Energy Component Average Std. Dev. Std. Err.
> of
> >> Mean
> >>
> >>
> -------------------------------------------------------------------------------
> >> VDWAALS 9.1993 4.5697
> >> 0.2408
> >> EEL -903.7460 17.4715
> >> 0.9208
> >> EGB 916.3956 14.1794
> >> 0.7473
> >> ESURF -2.3127 0.0888
> >> 0.0047
> >>
> >> DELTA G gas -894.5467 15.4122
> >> 0.8123
> >> DELTA G solv 914.0829 14.1770
> >> 0.7472
> >>
> >> DELTA TOTAL 19.5362 4.2174
> >> 0.2223
> >>
> >>
> ---------------------------------------------------------------------------------------
> >>
> >> PBSA:
> >>
> >> Differences (Complex - Receptor - Ligand):
> >> Energy Component Average Std. Dev. Std. Err.
> of
> >> Mean
> >>
> >>
> -------------------------------------------------------------------------------
> >> VDWAALS 9.1993 4.5697
> >> 0.2408
> >> EEL -903.7460 17.4715
> >> 0.9208
> >> EPB 868.5417 13.0425
> >> 0.6874
> >> ENPOLAR -11.2088 0.4629
> >> 0.0244
> >> EDISPER 19.7992 0.6073
> >> 0.0320
> >>
> >> DELTA G gas -894.5467 15.4122
> >> 0.8123
> >> DELTA G solv 877.1321 13.0279
> >> 0.6866
> >>
> >> DELTA TOTAL -17.4146 7.8453
> >> 0.4135
> >>
> >> We can see delta G of GBSA is 19.5362, where PBSA is -17.4146 and can be
> >> considered as a reasonable result for me.
> >> The polar interaction is dominant where EGB is 916.3956 and EPB is
> >> 868.5417 making the result become a positive and negative delta G,
> >> respectively.
> >>
> >> Obviously, GBSA overestimated the EGB, I want to know how can I do
> >> modification to parameter files to make GBSA's result become negative
> and
> >> reasonable??
> >>
> >> PS: I have tried IGB=5, but still the same
> >>
> >> Hope for your help.
> >> Chirs
> >> _______________________________________________
> >> 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
> _______________________________________________
> 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 Jun 18 2015 - 06:00:04 PDT
