[Genome] cool part 2

[Maximilian Haeussler]

Hi Tim,

I've come across this problem two or three years ago and I've simply
patched (yes, I know, a very very simple patch,but in this way I can
contribute something :-) a part of patmatch to directly output bed, so
you can  use it as-is, just import the bed track and play with it in
the table browser.
Put this into your patmatch directory ( I think I've changed four lines):
http://max.butterbrot.org/biofiles/scan_pipeline.pl

Robert: It would be nice if the browser would accept IUPAC codes,
should be a relatively quick patch as you already have code that is
searching for iupac. And yes, I also think that it would be nice if it
showed up as a normal track in the table browser.

This function could also search the mafs instead of the genome sequence.

While you're on it, there's code somewhere around ameme in the source
tree that is searching for "matrices", so you could also add it and
people could paste their matrices together with a cutoff so see where
they match... but this is just an idea...

And just thinking loud, again: I wonder if it wouldn't be possible one
day to use blat's index to search for these kind of short motifs
genome-wide... that would be a great tool with an unbeatable speed,
probably faster than those suffix tree solutions (see e.g. pmids
15933209 or 17067389). Perhaps it might already work out-of-box with
good parameters?

cheers,
Max

On 07/06/07, Robert Kuhn <kuhn at soe.ucsc.edu> wrote:
>
>
> this mailing-list question might also be useful.
> for degenerate codes, the full range of possible sequences
> would have to be included in the .fa file.
>
>
>  https://www.soe.ucsc.edu/pipermail/genome/2006-May/010707.html
>
>
>                         --b0b kuhn
>
>
> > From genome at soe.ucsc.edu  Thu Jun  7 14:28:48 2007
> > To: Tim Hefferon <thefferon at mail.nih.gov>
> > Cc: genome at soe.ucsc.edu
> > Subject: Re: [Genome] Fwd: cool part 2
> >
> > Good Afternoon Tim:
> >
> > Thanks for refreshing this important feature request.  We realize it
> > would be useful to have it behave like that.  Unfortunately it does
> > not at this time.
> >
> > At the current time we can only do such things off-line with the kent
> > source tree utilities.
> >
> > I have recently been investigating the ngrep program, along with
> > the patmatch front-end, as mentioned by Max in the genome email list:
> >       http://www.soe.ucsc.edu/pipermail/genome/2007-April/013327.html
> >       http://www.arabidopsis.org/cgi-bin/patmatch/nph-patmatch.pl
> >
> > --Hiram
> >
> > Tim Hefferon wrote:
> > > Begin forwarded message:
> > >
> > >> From: Tim Hefferon <timhefferon at gmail.com>
> > >> Date: April 22, 2007 9:19:05 AM GMT-06:00
> > >> To: genome at soe.ucsc.edu
> > >> Subject: cool part 2
> > >>
> > >> Hi Again,
> > >>
> > >> Crucial to the degenerate 'short match' feature I e-mailed about
> > >> earlier this morning would be the ability to access the data via
> > >> the Table Browser. Right now the 'short match' track is in the
> > >> Mapping and Sequencing Tracks section, but the data is not
> > >> available on the Table Browser (presumably because of the changing
> > >> nature of the track, depending on user input). Is there a way to
> > >> make the feature both user-changeable AND Table-Browsable? (might
> > >> this involve making the Short Match feature part of the Custom
> > >> Track field?)
> > >>
> > >> Thank you very much,
> > >> Tim
> > >>
> > >>
> > >> Timothy Hefferon, PhD
> > >> Research Fellow
> > >> Genome Technology Branch
> > >> National Human Genome Research Institute
> > >> National Institutes of Health
> > >> 50 South Dr., Rm. 5525
> > >> Bethesda, MD 20892-8004
> > _______________________________________________
> > Genome maillist  -  Genome at soe.ucsc.edu
> > http://www.soe.ucsc.edu/mailman/listinfo/genome
> >
> _______________________________________________
> Genome maillist  -  Genome at soe.ucsc.edu
> http://www.soe.ucsc.edu/mailman/listinfo/genome
>


-- 
Maximilian Haeussler,
skype: maximilianhaeussler