From chemistry-request@ccl.net Thu Dec  2 06:17:57 2004
Received: from mail.biosolveit.de (mail.biosolveit.de [194.8.222.62])
	by server.ccl.net (8.12.8/8.12.8) with ESMTP id iB2BHuYK008108
	for <CHEMISTRY~at~ccl.net>; Thu, 2 Dec 2004 06:17:56 -0500
Received: from localhost (localhost [127.0.0.1])
	by mail.biosolveit.de (Postfix) with ESMTP
	id 58C712CE67; Thu,  2 Dec 2004 12:32:48 +0100 (CET)
Received: from mail.biosolveit.de (localhost [127.0.0.1])
	by localhost (AvMailGate-2.0.1.16) id 11492-150FE280;
	Thu, 02 Dec 2004 12:32:48 +0100
Received: by mail.biosolveit.de (Postfix, from userid 65534)
	id 449642CE89; Thu,  2 Dec 2004 12:32:48 +0100 (CET)
Received: from lapsigma (iota.biosolveit.de [194.8.222.60])
	by mail.biosolveit.de (Postfix) with ESMTP
	id 73E532CE67; Thu,  2 Dec 2004 12:32:41 +0100 (CET)
From: =?iso-8859-1?Q?Holger_Clau=DFen?= <claussen~at~biosolveit.de>
To: <dirk.classen-houben~at~biovertis.com>, <CHEMISTRY~at~ccl.net>
Cc: "'Holger Claussen'" <hclaussen~at~biosolveit.de>
Subject: CCL:RE: FlexX und SGE
Date: Thu, 2 Dec 2004 12:32:21 +0100
Message-ID: <000f01c4d862$9b94ee80$2700000a@lapsigma>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
X-Priority: 3 (Normal)
X-MSMail-Priority: Normal
X-Mailer: Microsoft Outlook, Build 10.0.6626
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2180
Importance: Normal
In-Reply-To: <001a01c4d84c$47bdc630$1e00000a@lapalpha>
X-Spam-Status: No, score=-32.7 required=5.0 tests=ALL_TRUSTED,AWL,BAYES_00 
	autolearn=ham version=3.0.0
X-Spam-Level: 
X-Spam-Checker-Version: SpamAssassin 3.0.0 (2004-09-13) on dmz.biosolveit.loc
X-AntiVirus: checked by AntiVir MailGate (version: 2.0.1.16; AVE: 6.28.0.19; VDF: 6.28.0.101; host: dmz)
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by server.ccl.net id iB2BHvYK008109

Hello Dirk,

we recognized that it is actually a problem to combine 
FlexX and batch queuing systems (BQS) like for example 
SGE, because FlexX uses PVM for parallel computing. 
PVM set s up a separate control flow that conflicts 
with any BQS.

Thus we developed a small tool called Distributor that 
allows for setting up parallel FlexX jobs without PVM 
that employs a BQS, e.g. SEG.

The Distributor is a generic tool that can be applied 
to anycommand line based tool like for example FlexX. 
The only requirement is that the tool takes as input 
an ASCII file that contains the input for several runs 
of the tool and that this file can be split by rules 
(e.g. multi-mol2 ligand files for FlexX-Dockings).

The Distributor splits the input files and creates for
each batch a separate working environment and calls a 
job in the BQS to process the batch. The BQS is than 
responsible for the job scheduling. So there is only 
one control instance.
After the jobs are done, the Distributor gathers the 
ASCII output of all jobs and cats this to one output 
file automatically (e.g. the predicted scores of FlexX).

In addition, the Distributor can do some basic job 
statistic and report this optionally by email.
If the called tool (e.g. FlexX) is able to call a
callback script after each run, the distributor
is also able to restart failed jobs with the item 
following the failed one.

The split functions can be defined by the user.
Split functions for mol2, sdf, and for splitting 
ASCII files by pattern or by counting are pre defined. 
The Distributor works with several BQS. 
It is tested with SEG and OpenPBS.

If this sound interesting to you I would be happy
to send you more details on the Dirtsibutor.

Regards,
Holger


-- 
BioSolveIT GmbH - An der Ziegelei 75 - 53757 St.Augustin - Germany 
Dr. Holger Clau_en                          claussen~at~biosolveit.de 
Phone: +49-2241-2525-0 / Fax: -525               www.biosolveit.de


> -----Original Message-----
> From: Computational Chemistry List [mailto:chemistry-request~at~ccl.net] On
> Behalf Of Dirk Cla_en-Houben
> Sent: Monday, November 29, 2004 1:43 PM
> To: CCL
> Subject: CCL:FlexX und SGE
> 
> 
> Dear CCL readers,
> 
> has anybody some experience running FlexX from BiosolvIT and Tripos on a
> queuing system with the SGE (Sun Grid Engine)? Every suggestion or
> example is welkome.
> 
> Thanks in advance for your help,
> Regards,
> Dirk.
> 
> Dirk Classen-Houben
> BV Biotechnologies GmbH
> (Biovertis - information-driven drug design AG)
> Campus Vienna Biocenter 6
> 1030 Vienna, Austria
> 
> Phone +43-1-7989303-110
> Fax +43-1-7989303-400
> Email dirk.classen-houben@biovertis.c





