search

uec / Issue.Tracker

Automatically exported from code.google.com/p/usc-epigenome-center
0 stars 0 forks source link

Better ChIP-seq wiggle output #71

Open GoogleCodeExporter opened 8 years ago

GoogleCodeExporter commented 8 years ago 
The ChIP-seq wiggle output does not look publication quality. The reason is 
because it is only plotted at 50bp windows, which is not high enough 
resolution.  We could simply change this to 10 or 20bp windows, but I fear the 
resulting files might be too big. The other option is to have a data point at 
every base pair where a read starts or ends (this is similar to Peggy Farnham's 
sole-search output).  I put a copy of the Stanford code Peggy's group uses here:
~bberman/bin/noisySgrWigMF.py

Another option would be to update and install Sole-search.  Another would be to 
use FindPeaks.  Unfortunately, the FindPeaks wig files that are currently 
exported are not ideal because they have non-peak reads removed, which we don't 
want.  But maybe FindPeaks can export a pre-subtraction wig.

Original issue reported on code.google.com by benb...@gmail.com on 3 May 2011 at 3:42

GoogleCodeExporter commented 8 years ago 
superseded by bindepths wig/bigwig generation

Original comment by zack...@gmail.com on 8 Nov 2012 at 10:44

GoogleCodeExporter commented 8 years ago

Original comment by zack...@gmail.com on 8 Nov 2012 at 10:44

GoogleCodeExporter commented 8 years ago 
I want to keep this issue open, because I don't think our "bindepth30" is the 
best solution for ChIP-seq.  Charlie is right that in order to really make the 
best density plot for ChIP-seq, you need to know the fragment length.  But you 
really should do read extension, see attached illustrations of why.  There's 
also the simple SoleSearch script listed above.

I am assigning this issue to myself even though I may not have time to get to 
it for a while.  Maybe Lijing can help with this since she's worked a lot with 
ChIP-seq.

Original comment by benb...@gmail.com on 8 Nov 2012 at 11:12

Attachments:

GoogleCodeExporter commented 8 years ago 
Also see issue #191 for information how to use FindPeaks for this:
http://code.google.com/p/usc-epigenome-center/issues/detail?id=191

Original comment by benb...@gmail.com on 8 Nov 2012 at 11:15