ADL: MPI.Autodock4

Collignon, Barbara C. collignonbc at ornl.gov
Fri Apr 17 15:55:04 PDT 2009


ok. I well know the first paper. I got the MPI version few months ago
but we had problems to install it. As well, nobody (from the paper) 
help to resolve the bug.

Barbara 


-----Original Message-----
From: autodock-bounces at scripps.edu on behalf of Mark Swingle
Sent: Fri 4/17/2009 6:38 PM
To: autodock at scripps.edu
Subject: Re: ADL: MPI.Autodock4
 
Sorry Barbara, I wasn't paying enough attention.

I'm not aware of any download links. I suppose one would have to contact the 
authors of this paper (not sure if they modified AD3 or AD4, though):
 http://scripts.iucr.org/cgi-bin/paper?S0021889807011053

I also recall someone posting here last year who was working on an MPI 
version:

http://n2.nabble.com/ADL:-mpi-programming-issue-with-autodock4-and-ligands-with-0-active-torsions-td1496995.html

I think a few people on the list have managed to acquire one of these two 
versions. Maybe one of them will be willing to send it to you if you want it.

Regards,

Mark


On Friday 17 April 2009 04:57:40 pm Collignon, Barbara C. wrote:
> thanks Mark, but I already know that (see my previous messages).
> I just wanted to know where one can download this famous MPI.Autodock4.
>
> Barbara
>
> -----Original Message-----
> From: autodock-bounces at scripps.edu on behalf of Mark Swingle
> Sent: Fri 4/17/2009 4:36 PM
> To: autodock at scripps.edu
> Subject: Re: ADL: MPI.Autodock4
>
> Barbara,
>
> For virtual screening, an MPI version of autodock is not needed. VS is what
> one might call embarassingly parallel. You simply need to divide your
> ligand database into small segments and run autodock processes with a
> segment on each available processor core. This can be done with simple
> scripting (numerous examples are floating around) or with something like
> DOVIS that is specifically geared to VS. My guess is that multiple
> instances of "plain vanilla" autodock running in parallel on many
> processors would be a bit faster than a single instance of an MPI version
> of autodock utilizing the same number of processors, simply because of the
> increased overhead due to MPI.
>
> Regards,
>
> Mark
>
> On Friday 17 April 2009 09:34:51 am Collignon, Barbara C. wrote:
> > For VS
> >
> > Barbara
> >
> >
> > -----Original Message-----
> > From: autodock-bounces at scripps.edu on behalf of M. Shahid
> > Sent: Fri 4/17/2009 2:59 AM
> > To: autodock at scripps.edu
> > Subject: Re: ADL: MPI.Autodock4
> >
> > Do you need it for single run on how many processors? or you need one for
> > VS?
> >
> >
> > On Thu, Apr 16, 2009 at 10:52 PM, Collignon, Barbara C. <
> >
> > collignonbc at ornl.gov> wrote:
> > > Sorry , I am confused.
> > >
> > > May someone give me the link to download MPI.Autodock4 ?
> > >
> > > Barbara
> > >
> > >
> > >
> > > ________________________________________________
> > > --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list
> > > ---
> >
> > ________________________________________________
> > --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---
> >
> >
> > ________________________________________________
> > --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---
>
> ________________________________________________
> --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---
>
>
> ________________________________________________
> --- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---


________________________________________________
--- ADL: AutoDock List  --- http://autodock.scripps.edu/mailing_list ---




More information about the autodock mailing list