Also for the record. I've found that splitting a database in this way has the extremely beneficial side effect of reducing the total RAM required by BLAST (over blasting against a single database of the exact same sequences). It seems to scale with the number of databases created. Though I've not verified with any rigorous testing, so reader beware. ----- Original Message ----- From: "Wim Glassee" <wim.glassee@ua.ac.be> To: <bioclusters@bioinformatics.org> Sent: Tuesday, January 21, 2003 12:54 AM Subject: RE: [Bioclusters] Details on a local blast cluster question Hi Joe, Just for the record, In the new version of formatdb the -v parameter takes MB's in stead of bytes. So the command would be: formatdb -v 100 -p T -o T -i nr > -----Original Message----- > From: bioclusters-admin@bioinformatics.org [mailto:bioclusters- > admin@bioinformatics.org] On Behalf Of Joe Landman > Sent: maandag 20 januari 2003 23:57 > To: bioclusters@bioinformatics.org > Subject: Re: [Bioclusters] Details on a local blast cluster question > > Hi Tracy: > > Look at the -v switch on formatdb. The -v switch breaks the database > into volumes and preserves the complete database size (so you do not > have to recalculate the size-dependent quantities upon completion). > > For exanple, to split a recent nr into several volumes of at most 100M > letters > > formatdb -v 100000000 -p T -o T -i nr > > yields > > [root@frontend-0 db2]# ls -alF > total 1507984 > drwxr-xr-x 2 root root 4096 Jan 20 22:53 ./ > drwxr-xr-x 10 root root 4096 Jan 20 22:43 ../ > -rw-r--r-- 1 root root 405 Jan 20 22:53 > formatdb.log > -rw-r--r-- 1 root root 589765675 Jan 20 22:44 nr > -rw-r--r-- 1 root root 190609908 Jan 20 22:49 nr.00.phr > -rw-r--r-- 1 root root 6949984 Jan 20 22:51 nr.00.pin > -rw-r--r-- 1 root root 12565120 Jan 20 22:49 nr.00.pnd > -rw-r--r-- 1 root root 49132 Jan 20 22:49 nr.00.pni > -rw-r--r-- 1 root root 173748036 Jan 20 22:50 nr.00.psd > -rw-r--r-- 1 root root 3804603 Jan 20 22:51 nr.00.psi > -rw-r--r-- 1 root root 277315926 Jan 20 22:49 nr.00.psq > -rw-r--r-- 1 root root 82065838 Jan 20 22:53 nr.01.phr > -rw-r--r-- 1 root root 2997568 Jan 20 22:53 nr.01.pin > -rw-r--r-- 1 root root 5409488 Jan 20 22:53 nr.01.pnd > -rw-r--r-- 1 root root 21180 Jan 20 22:53 nr.01.pni > -rw-r--r-- 1 root root 74213127 Jan 20 22:53 nr.01.psd > -rw-r--r-- 1 root root 1658969 Jan 20 22:53 nr.01.psi > -rw-r--r-- 1 root root 119641224 Jan 20 22:53 nr.01.psq > -rw-r--r-- 1 root root 106 Jan 20 22:53 nr.pal > > To run the code, simply use the normal methods. Use nr as the database > (-d nr) > > Joe > > -- > Joseph Landman, Ph.D > Scalable Informatics LLC, > email: landman@scalableinformatics.com > web : http://scalableinformatics.com > phone: +1 734 612 4615 > > _______________________________________________ > Bioclusters maillist - Bioclusters@bioinformatics.org > https://bioinformatics.org/mailman/listinfo/bioclusters _______________________________________________ Bioclusters maillist - Bioclusters@bioinformatics.org https://bioinformatics.org/mailman/listinfo/bioclusters