We recommend the -fastMap option for large chunks
that should align in a single piece. This will
make it run much faster. Otherwise there is a
finishing step that is order n-squared running time
with query-size.
We found that when the query size exceeded 5000 bases,
on rare occasions we missed a small piece of the alignment.
As a work-around we have restricted the maximum size
of the query to 5000 bases.
#define MAXSINGLEPIECESIZE 5000 /* maximum size of a single piece */
if (fastMap && (seq->size > MAXSINGLEPIECESIZE))
errAbort("Maximum single piece size (%d) exceeded by query %s of
size (%d). "
"Larger pieces will have to be split up until no larger than
this limit "
"when the -fastMap option is used."
, MAXSINGLEPIECESIZE, seq->name, seq->size);
Without -fastMap option, it will run with larger queries,
but it may be very slow.
When we are running large blat alignments,
we tend to break the queries into overlapping pieces
and we can run various blat jobs a cluster to speed it up more
and then use a chaining program to rejoin broken alignments
at the end.
-Galt
7/3/2011 11:53 AM, Young, Lynn (NIH/OD/ORS) [E]:
> Greetings,
>
> Regarding the FAQ section "BLAT limitations".
It says "DNA input sequences are limited to a maximum length of 25,000
bases.
Protein or translated input sequences must not exceed 5000 letters."
>
> Is this true for command line BLAT, as well?
>
> Sincerely,
> Lynn
>
> Lynn Young, PhD | Informationist/Bioinformatics | NIH Library | National
> Institutes of Health | Building 10, Room B1L-306D | 10 Center Drive |
> Bethesda, Maryland 20892-1150 | T. 301-594-6426 | F. 301-402-0254 | E.
> [email protected]<mailto:[email protected]> |
> http://nihlibrary.nih.gov<http://nihlibrary.nih.gov/> |
> http://nihlibrary.nih.gov/Bioinformatics | http://twitter.com/bioinformatics1
> _____________________________
> Amazing Research. Amazing Help
>
>
> _______________________________________________
> Genome maillist - [email protected]
> https://lists.soe.ucsc.edu/mailman/listinfo/genome
_______________________________________________
Genome maillist - [email protected]
https://lists.soe.ucsc.edu/mailman/listinfo/genome
