I think the idea for -fastMap is that it doesn't
try to chain alignments together, so no introns.

The "high %ID" required for said fastMap is not modifiable.
Presumably it's very high (>90%).

Of course simple %-identity threshold filtering is set with -minIdentity.

fastMap does not enhance sensitivity but it might
be good for speed in situations where gaps are
not expected.

I see that it is used for liftOver from one assembly
to the next of the same species;
and for mapping clones to contigs;

Essentially any place where you expect really long
identical alignments with no gaps or introns.

Normally blat runs a banded dynamic programming algorithm
to finalize and extend alignments.  This is an expensive
and slow step and does not scale well with long queries.

fastMap helps blat perform better when there are these
long identical regions by skipping those extra costly slow
steps.  Even so, applications usually chunk the query
into 3kb pieces and run cluster jobs.
Blat/gfServer have a default limit query size of 40kb.
It just runs too slowly for longer queries.

Several examples of fastMap usage look like this:
  blat -minScore=100 -minIdentity=98 -fastMap

fastMap is for mapping a chunk of virtually identical
dna onto a larger object like a clone or scaffold or chrom.

It will tolerate a few substitutions but not much else.
Certainly no long gaps.

-Galt


On Wed, 25 Mar 2009, Royden Clark wrote:

> Greetings UCSC Genome Group,
>
> Sorry if this question has already been answered but I could not find
> it in the archives or the faqs.
>
> The documentation for blat
> at
> http://genome.ucsc.edu/goldenPath/help/blatSpec.html
>
> states
>
>  -fastMap    Run for fast DNA/DNA remapping - not allowing introns,
>
>                requiring high %ID
>
>
>
>
> The question is what is the %ID it requires?
> Can that % be modified?
>
> Thank you
> Royden Clark
>
> IT Specialist II
> University of Virginia
> _______________________________________________
> 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