FYI: pbsa in AT17 does have CUDA-enabled. (But I am not sure how/if you
can use that with mm-pbsa)
Hai
On Wed, Jun 28, 2017 at 12:00 AM, Hai Nguyen <nhai.qn.gmail.com> wrote:
>
>
> On Tue, Jun 27, 2017 at 11:41 PM, Baker, Joseph <bakerj.tcnj.edu> wrote:
>
>> Hai,
>>
>> We are also running with mm-pbsa. The calculations with mm-pbsa are taking
>> a bit (one of them seemed to stall after a couple of frames, another we
>> need to increase maxitn it seems), so while those are running we've
>> decided
>> to do some gbsa tests as well to compare. We are using ambertools 17. I'll
>> go ahead and run that test to make sure it checks out.
>
>
>
>
>> Would you recommend
>> non linear PB for protein DNA by the way?
>>
>
> hi, Ray would be the best person to answer this question.
> btw, I am very curious about the comparison between PB and igb8 in your
> case once you're done.
>
> cheers
> Hai
>
>
>>
>> Joe
>>
>> -Sent from Android mobile
>>
>> On Jun 27, 2017 11:27 PM, "Hai Nguyen" <nhai.qn.gmail.com> wrote:
>>
>> > Hi,
>> >
>> > why don't you use mm-pbsa? mm-gbsa is not *that* accurate and I am
>> doubting
>> > you will get meaningful dG.
>> >
>> > per igb=8, which ambertools version you are using? Did this tests pass
>> for
>> > sander?
>> >
>> > cd $AMBERHOME/test/gbneck2nu && make
>> >
>> > Hai
>> >
>> > On Tue, Jun 27, 2017 at 11:18 PM, Baker, Joseph <bakerj.tcnj.edu>
>> wrote:
>> >
>> > > Hi all,
>> > >
>> > > We're trying to do some tests to compare mm-pbsa and mm-gbsa
>> calculations
>> > > for a protein/DNA complex. We've now got our mm-pbsa calculations
>> mostly
>> > > working. For the mm-gbsa calculation, in our &gb section of the
>> > mmgbsa.in
>> > > script to MMPBSA.py we are trying igb=5 and igb=8 (both with mbondi2
>> for
>> > > the radius set, though we've also tried mbondi3 for the igb=8 case).
>> In
>> > the
>> > > igb=5 case we are getting a negative dG, while using igb=8 with either
>> > > mbondi2 or mbondi3 we are getting very positive dG values. I've seen
>> some
>> > > previous discussions here about folks having some difficulty with
>> mm-gbsa
>> > > for protein/nucleic complexes, but it seemed that those were prior to
>> the
>> > > optimized nucleic gb parameters of igb=8. Can anyone provide any
>> input on
>> > > whether mm-gbsa is a reasonable approach for relative dG's with
>> either of
>> > > these GB models? Thanks for your input.
>> > >
>> > > Kind regards,
>> > > Joe
>> > > _______________________________________________
>> > > 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 Tue Jun 27 2017 - 21:30:03 PDT
