It looks like there are some repetitive tiles in your
sequence that make it harder to detect.

If you add -repMatch=2048 to standalone blat commandline,
you will get the hit you wanted.

You could increase it even more and find other lower-quality
alignments.

I took the default repMatch of 1024 and doubled it because
you are using stepSize=5.

-Galt


On Thu, 12 Mar 2009, Amol Carl Shetty wrote:

> Hello
>
> I have been using the standalone Blat extensively over the last few
> months using the following command :
>
> blat -stepSize=5 -minScore=0 -minIdentity=0 database.2bit query.fa
> output.psl
>
> which replicates the web-based version of the same as specified on the
> website (http://genome.ucsc.edu/FAQ/FAQblat).
>
>
> However, I have run into an issue where there is a significant
> difference in the results. For example,
>
> On using the web-based Blat based on the following :
>
> Genome                : Mouse
> Assembly      : July 2007
> Query Type    : DNA
> Sequence      :
> ATTTGTGTGTGTGTGTGTGTTTTTTTTTTTGTTGTTGTTGTTGTTGTTGTTTTGAGACAGGGTTT
>
> The result obtained is
> QUERY         SCORE           START           END     QSIZE   IDENTITY        
> CHRO    STRAND  START           END             SPAN
> ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------
> YourSeq           65                        1                    65           
>      65
> 100.0%             16         +               62887839        62887903        
>     65
>
>
> However, on using the standalone blast, the same sequence yields no
> results.
>
> Kindly, let me know where the problem lies?
>
> Thank you
>
> Amol Carl Shetty
> Department of Human Genetics
> Emory University
> Atlanta, Georgia
> _______________________________________________
> Genome maillist  -  [email protected]
> http://www.soe.ucsc.edu/mailman/listinfo/genome
>
_______________________________________________
Genome maillist  -  [email protected]
http://www.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to