I have reproduced this behavior.
The sequence is not unique despite its length of 27, and maps to at least hundreds of places on the genome. This location is in a LINE repeat at the end of chr1. This is the known behavior of the over-used tile with blat, where blat stops adding locations to the tile index once it's reached a maximum number. I tried to increase -repMatch to a large number, but it did not help. Then I extracted a subset of chr1, keeping only bases after 200Million. Only the last 47 million were left. I ran blat against this portion of chr1. This worked fine and turned up 32 hits including the one that you were expecting. It is difficult to index the entire genome to its full depth when some times would be hit millions of times. Handling repeats exhaustively is not one of BLAT's goals. Jim Kent is working on a new short-sequence aligner program that might help here, but it is still in development. Repeat RepeatMasker Information Name: L1PA7 Family: L1 Class: LINE SW Score: 16289 Divergence: 8.7% Deletions: 0.8% Insertions: 0.8% Begin in repeat: 3132 End in repeat: 4244 Left in repeat: 1902 Position: chr1:247183181-247184298 Band: 1q44 Genomic Size: 1118 Strand: - -Galt On Tue, 4 Nov 2008, Yuan Jian wrote:
Hi there, I got a seuquence TGAAAACTGGCACAAGACAAGGATGCC located in chr1:247183338-247183364 strand-. but when I blat it. I can not find that location for the sequence. but I found it in other loci of chr1 and strand-: browser details YourSeq 27 1 27 27 100.0% 1 - 219139486 219139512 27 browser details YourSeq 27 1 27 27 100.0% 1 - 192113590 192113616 27 browser details YourSeq 27 1 27 27 100.0% 1 - 189295101 189295127 27 browser details YourSeq 27 1 27 27 100.0% 1 - 156553131 156553157 27 browser details YourSeq 27 1 27 27 100.0% 1 - 156665481 156665507 27 can you please tell me why? thanks Yu _______________________________________________ Genome maillist - [email protected] http://www.soe.ucsc.edu/mailman/listinfo/genome
_______________________________________________ Genome maillist - [email protected] http://www.soe.ucsc.edu/mailman/listinfo/genome
