AMBER Archive (2009)

Subject: Re: [AMBER] pairwise per-residue decomposition

From: manoj singh (mks.amber_at_gmail.com)
Date: Sat Oct 31 2009 - 23:59:38 CDT

Hi all !

 Just to conclude here, I got the pairwise per-residue decomposition working
without using mm_pbsa.pl. I needed to write few extra script and there is no
memory issue now.

Thanks for all the suggestions.

Manoj

On Fri, Oct 30, 2009 at 2:26 PM, manoj singh <mks.amber_at_gmail.com> wrote:

> Thanks for the reply.
>
> I think problem was something else.
>
> In the sander output, there are five terms; internal, vdw, eel, Gpol, ans
> sas. I think sas is the solvent accessible surface are which need to
> multiply by 0.0072 to get the Gsa.
>
> Am I right?
>
> I will be very thankful for the reply.
>
> Manoj
>
>
> On Fri, Oct 30, 2009 at 2:18 PM, Tom Joseph <ttjoseph_at_gmail.com> wrote:
>
>> That is possible...presumably the other residues cancel out its
>> effect. Also, did you average over a number of frames?
>>
>> --Tom
>>
>> 2009/10/30 manoj singh <mks.amber_at_gmail.com>:
>> > Thanks for your reply.
>> >
>> > I got this part working.
>> >
>> > Now, I want to calculated the contribution of the residue towards ligand
>> > binding affinity( the ligand is a peptide). I am adding binding affinity
>> > contribution of that residue of protein with all residues of the ligand.
>> > However, the number I am getting is higher than the total MM-GBSA
>> binding
>> > affinity of the ligand.
>> >
>> > I will be very thankful for any suggestion.
>> >
>> > Sincerely,
>> > Manoj
>> >
>> > On Fri, Oct 30, 2009 at 1:15 AM, Tom Joseph <ttjoseph_at_gmail.com> wrote:
>> >
>> >> To get an energy decomposition, run sander with imin=1 and idecomp set
>> >> to the type of analysis you want (in this case 4). Run with no
>> >> periodic boundary conditions (ntb=0) and an "infinite" cutoff
>> >> (cut=999). See below for an example mdin file that does an MM/GBSA
>> >> calculation on a 709-residue structure. It will generate a huge mdout
>> >> file which you can then postprocess manually as you wish. Please be
>> >> sure to check the values of the parameters to be sure they are what
>> >> you really want before using them.
>> >>
>> >> --Tom
>> >>
>> >> complex.crd.1 (MM)
>> >> &cntrl
>> >> cut = 999.0,
>> >> dielc = 1.0,
>> >> extdiel = 80.0,
>> >> gbsa = 2,
>> >> idecomp = 4,
>> >> igb = 2,
>> >> imin = 1,
>> >> intdiel = 1.0,
>> >> maxcyc = 1,
>> >> ncyc = 0,
>> >> nsnb = 99999,
>> >> ntb = 0,
>> >> ntf = 1,
>> >> offset = 0.09,
>> >> saltcon = 0.0,
>> >> scee = 1.2,
>> >> scnb = 2.0,
>> >> surften = 0.0072,
>> >> &end
>> >> Residues for decomposition
>> >> LRES 1 709
>> >> END
>> >> Residues to print
>> >> RES 1 709
>> >> END
>> >> END
>> >>
>> >>
>> >> 2009/10/29 manoj singh <mks.amber_at_gmail.com>:
>> >> > Thanks for your reply!
>> >> > I relatively new to Amber and therefore not very experienced in doing
>> >> these
>> >> > type of calculations. I will be very thankful if you can please tell
>> me
>> >> how
>> >> > exactly did yo perform this calculation without using mm_pbsa.pl. I
>> am
>> >> > trying to do this calculation for last 3 weeks but has not
>> >> been successful.
>> >> > I will be very thankful for your kind reply.
>> >> > Sincerely,
>> >> > Manoj
>> >> > On Thu, Oct 29, 2009 at 8:30 PM, Tom Joseph <ttjoseph_at_gmail.com>
>> wrote:
>> >> >>
>> >> >> I had the same problem with this script for ~1000 frames of my ~700
>> >> >> residue system on a machine with 8GB of RAM. It would just sit
>> there.
>> >> >> I don't know what the problem is, but I gave up on it and now just
>> run
>> >> >> sander manually and postprocess its output outside of MM_PBSA.pl.
>> >> >>
>> >> >> You can try running "top" in another terminal session on the same
>> >> >> machine while your 1000-frame calculation is attempting to run to
>> get
>> >> >> an idea of its memory usage. Look at the "RES", "RSS", or similar
>> >> >> column.
>> >> >>
>> >> >> --Tom
>> >> >>
>> >> >> 2009/10/29 manoj singh <mks.amber_at_gmail.com>:
>> >> >> > Hi all,
>> >> >> >
>> >> >> > I am trying to do pairwise per-residue decomposition of MM-GBSA
>> >> binding
>> >> >> > free
>> >> >> > energy for a protein ligand system. My calculation is crashing
>> (the
>> >> >> > computer
>> >> >> > stop responding) on the "=>> Calc delta from raw data", probably
>> due
>> >> to
>> >> >> > the
>> >> >> > memory issue. I will be very thankful if some one can tell me a
>> fix of
>> >> >> > this
>> >> >> > problem.
>> >> >> >
>> >> >> > Manoj
>> >> >> > _______________________________________________
>> >> >> > AMBER mailing list
>> >> >> > AMBER_at_ambermd.org
>> >> >> > http://lists.ambermd.org/mailman/listinfo/amber
>> >> >> >
>> >> >>
>> >> >> _______________________________________________
>> >> >> AMBER mailing list
>> >> >> AMBER_at_ambermd.org
>> >> >> http://lists.ambermd.org/mailman/listinfo/amber
>> >> >
>> >> >
>> >>
>> >> _______________________________________________
>> >> AMBER mailing list
>> >> AMBER_at_ambermd.org
>> >> http://lists.ambermd.org/mailman/listinfo/amber
>> >>
>> > _______________________________________________
>> > AMBER mailing list
>> > AMBER_at_ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber
>> >
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER_at_ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
>
_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber