AMBER Archive (2009)

Subject: Re: [AMBER] MD snapshots of most probable protein conformations in Amber

From: Andrew Voronkov (drugdesign_at_yandex.ru)
Date: Mon Nov 30 2009 - 06:56:42 CST

I have found the workflow which I was looking for, but some questions on ptraj usage for clustering remains.

The ptraj clustering command format is listed here for your edification:

cluster out filename [representative format] [average format] [all format] algorithm ([clusters n] | [epsilon critical_distance])
[rms|dme] [sieve s [start start_frame | random]] [verbose verb] [mass] mask

then the exact ptraj file is listed:

"* Start with cluster=1
trajin N1-apo-all-CA-aligned.dcd
cluster out avg-1 representative none average none all none averagelinkage clusters 1 mass rms
 "(:35-37 | :51-56 | :64-70 | :74 | :97-98 | :114-118 | :141-146 | :161-165 | :195-196 | :211 | :213 | :262-265 | :286 | :319-320 |
:344-359) & @CA,C,N,O"
go"

In my case I have made the next ptraj file:

trajin fzd2_w_md5.mdcrd
cluster out avg-1 representative none average none all none averagelinkage clusters 1 mass rms
 "(:1-240) & @CA,C,N,O"
go

but I don't know what file to specify as output? .rst, .pdb or .mdcrd formats for clustered conormations don't work I
also wonder about more detailed meaning of the flags, like the representative structure - this is representative word after cluster command?

I use then ptraj protein.prmtop cluster .ptraj ?output file??

I haven't found information on this in Amber10 or Amber9 manuals.

Best regards,
Andrew

18.11.09, 12:14, "Carlos Simmerling" <carlos.simmerling_at_gmail.com>:

> one could do that, but most clustering algorithms have a built-in method for
> selecting the "representative" structure for the cluster.
>
> On Wed, Nov 18, 2009 at 10:32 AM, Andrew Voronkov wrote:
>
> > So I get the conformations clustered and then just select for example the
> > conformations in each cluster with lowest energy, right?
> >
> > Best regards,
> > Andrey
> >
> > 18.11.09, 08:53, "Carlos Simmerling" :
> >
> > > the ptraj program that Andrew asked about also supports clustering, so if
> > he
> > > already knows how to use ptraj there's no reason to learn MMTSB.
> > > On Wed, Nov 18, 2009 at 8:39 AM, Renata KWIECIEN <
> > > Renata.Kwiecien_at_univ-nantes.fr> wrote:
> > > > Check "cluster" option in MMTSB.
> > > > http://blue11.bch.msu.edu/mmtsb/Main_Page
> > > >
> > > > Best regards,
> > > > Renata
> > > >
> > > > > Actually the question is about how to generate protein conformations
> > > > > ensemble and then to select the most probable conformations.
> > > > > It s possible to generate receptors ensemble using namd and it should
> > be
> > > > > possible by using amber, but I need let s say 30 most probable
> > > > > conformations of proteins which have the highest occurence rate in
> > the
> > > > > 5-10 nanosecond trajectory.
> > > > >
> > > > > Is it possible to do it by ptraj for example?
> > > > >
> > > > > Best regards,
> > > > > Andrew
> > > > >
> > > > > 12.11.09, 08:30, "Adrian Roitberg" :
> > > > >
> > > > >> I am not sure, but I believe that Andy Mc Cammon's group has
> > implemented
> > > > >> this setup in their own web site, extending it from namd as in the
> > paper
> > > > >> you mentioned, to amber now.
> > > > >> Adrian
> > > > >> Andrew Voronkov wrote:
> > > > >> > Dear Amber users,
> > > > >> > are there any tutorials online on usage of some analogs of relaxed
> > > > >> complex scheme, described here:
> > > > >> >
> > > >
> > http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2516539/?tool=pubmed#aff-info
> > > > >> >
> > > > >> > I mean optimization of MD trajectory snapshots and usage of most
> > > > >> probable protein conformations from MD trajectories in Amber?
> > > > >> >
> > > > >> >
> > > > >> > Best regards,
> > > > >> > Andrey
> > > > >> >
> > > > >> > _______________________________________________
> > > > >> > 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
>
>

_______________________________________________
AMBER mailing list
AMBER_at_ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber