oops, I meant to say .ooc and not .00c. -Galt
On Fri, 20 Feb 2009, Galt Barber wrote: > > I might add that when running one chrom at a time with > blat you can help the consistency of your results by > making the .00c file and then using it with each > chrom blat command. > > -Galt > > > On Fri, 20 Feb 2009, Hiram Clawson wrote: > >> Yes, run one chromosome at a time. The in-core image of >> blat when reading the entire genome and making the ooc index >> is immense. On the order of 6 to 8 Gb of memory depending >> upon stepSize. >> >> --Hiram >> >> Doug Wolfgram wrote: >>> I am having the same problem as Helder in a thread from last spring. I >>> have a Centos 5.2 box with 4GB of ram, but when I try to BLAT a small >>> sequence: >>> >>> CATGCCATCCTGCGTCTGGACCTGGCTGGCZGGGACCTGACTGACTACCTCATGAAGATCC (Z is >>> the muation target) >>> >>> I get the out of memory error. >>> >>> needLargeMem: Out of memory - request size 154913755 bytes >>> >>> Shouldn't this work on a 4GB memory machine? Is the only solution to >>> write a loop in the bash script to check one chromosome at a time? >>> >>> One other thing. How do I get a 120mer output? I am inputting 60mer >>> sequences and need to extend them to 120. >>> >>> ------------- >>> Doug Wolfgram >>> [email protected] >>> >>> >>> >>> _______________________________________________ >>> Genome maillist - [email protected] >>> http://www.soe.ucsc.edu/mailman/listinfo/genome >>> >> >> _______________________________________________ >> Genome maillist - [email protected] >> http://www.soe.ucsc.edu/mailman/listinfo/genome >> > _______________________________________________ > Genome maillist - [email protected] > http://www.soe.ucsc.edu/mailman/listinfo/genome > _______________________________________________ Genome maillist - [email protected] http://www.soe.ucsc.edu/mailman/listinfo/genome
