Hi, Louis! Problem #2 is a coordinate bug in webBlat that will be fixed soon hopefully. I have notified Jim Kent the author of BLAT about it.
Problem #1 about the differing join threshold of 8 hgBlat versus 20 webBlat is something that I can also raise with Jim in the interests of consistency. Thanks for reporting these issues. In future, please direct questions about BLAT to the main mailing list "genome" instead of "genome-mirror". -Galt On Mon, 22 Dec 2008, louis wrote: > Hi, > > I try to replicate a UCSC blat service on my own machine. > I start gfServer with the following parameter, as specified in the Blat > FAQ (http://genome.ucsc.edu/FAQ/FAQblat): > > gfServer start blatMachine portX -stepSize=5 -log=untrans.log > database.2bit > > > And I access gfServer with webBlat CGI. > > However, there are two problems: > 1. BLAT Search Results from my own blat service are not always exactly > the same as that from the UCSC blat service. > > * My blat result: > > query 800 1 800 800 100.0% > gi|89161185|ref|NC_000001.9|NC_000001 + 101997551 101998350 800 > query 30 330 361 800 96.9% > gi|89161210|ref|NC_000006.10|NC_000006 + 51189132 51189163 32 > query 28 332 359 800 100.0% > gi|89161216|ref|NC_000009.10|NC_000009 - 124137815 124137842 28 > query 27 625 669 800 93.6% > gi|51511721|ref|NC_000005.8|NC_000005 + 14649289 14649656 368 > query 27 334 360 800 100.0% > gi|89161213|ref|NC_000007.12|NC_000007 + 85422415 85422441 27 > > > * UCSC blat Result: > > browser details YourSeq 800 1 800 800 100.0% > 1 + 101997551 101998350 800 > browser details YourSeq 33 325 358 800 100.0% > 13 - 59690766 59690802 37 > browser details YourSeq 30 330 361 800 96.9% > 6 + 51189132 51189163 32 > browser details YourSeq 28 332 359 800 100.0% > 9 - 124137815 124137842 28 > browser details YourSeq 27 334 360 800 100.0% > 7 + 85422415 85422441 27 > > > My Blat instance generated an extra record: > > query 27 625 669 800 93.6% > gi|51511721|ref|NC_000005.8|NC_000005 + 14649289 14649656 368 > > > I assume it's something relevant to the gap-merging issue since that > extra record has gaps with the size between 8 and 20 bases, and my Blat > instance shows "//Aligned Blocks with gaps <= 20 bases are merged for > this display" ; //while what of UCSC Blat shows "/Aligned Blocks with > gaps <= 8 bases are merged for this display /" However, I can't find > out how to adjust this parameter. > > > 2. for the details of each Blat result listed in the Blat Search Result, > the start and end position are offset by 20 bases: > > * the start and end shown in the Blat Search Result: 101997551 - > 101998350 > > query 800 1 800 800 100.0% > gi|89161185|ref|NC_000001.9|NC_000001 + 101997551 101998350 800 > > > > * That shown in the corresponding details is: 101997671 - 101998470 > > 000000001 > tgttactgatagtaaattcttaaatatattctgtgctggtttactggttcattatttgaa 000000060 > >>>>>>>>> > |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| >>>>>>>>> > 101997671 > tgttactgatagtaaattcttaaatatattctgtgctggtttactggttcattatttgaa 101997730 > ...... > 000000781 tcatcatcctcaaatgatta 000000800 > >>>>>>>>> |||||||||||||||||||| >>>>>>>>> > 101998451 tcatcatcctcaaatgatta 101998470 > > The spec to produced this result is as follows: > > * OS: CentOS kernel 2.6.18 > * gfServer/webBlat : blatSuite v34 > * Apache 2 > * Human Genome assembly build 36.1 > * Query sequence: > > tgttactgatagtaaattcttaaatatattctgtgctggtttactggttc > attatttgaattcttcttgattggtttatttgtgaaaatatatttatttc > gcctacatttttgaaggctattttatatgagtataggaatgtaagttggc > agttacttctcttagtacgttaaagattttgttccatttttggttggatt > ctgttattctcttgagaaatcagttgtaaaattaatttttgttactttga > gtaatgtgccttttaaaattttattccctgtttccagattctctcctttt > gtatgagattttcttctgtaatgtgattaggtgtgtttttgttgttgttg > ttattgtttttattgcatttaaattttccagtgcctcttgaatatgtgga > ttgttgtctttcctgagttctggattctcccctgtcttctcttcagattt > tgcttctgcttcatttcctttctGATTTGTGTTTGTGTTTGGGGCTATAT > CTACATTTATGTCAGATATTTGACTGTGTTACACACAGCTCCTTTTTTTT > CAACTTTCCACTTTTTTAGATGTGCTTTTGTTTACATATTTTGACTATCT > GCTATTAAATTGTTTAAATTCTCAACACTCCTTTTTTTTCATGATATTTT > TCTTTTGGGATAATTTTTTCCCTGACAATAAAAATAAAAAGtttcctgaa > acattaaaaaatacatttagtgagggtttttttaaatgattaaacactta > cagtttgtatttctacatatttaaaacttatcatcatcctcaaatgatta > > > > > Cheers, > > Louis > > > > > > _______________________________________________ > Genome-mirror mailing list > [email protected] > http://www.soe.ucsc.edu/mailman/listinfo/genome-mirror > _______________________________________________ Genome maillist - [email protected] http://www.soe.ucsc.edu/mailman/listinfo/genome
