Hello Dr. Telgenkamp,

Your posting was forwarded to me by a colleague, and I felt it would be 
appropriate to reply to it, despite not being a regular user of this 
mailing list, as I currently maintain and develop the FindPeaks software 
package.

I have seen the problem you're describing before, and it appears to have 
a relatively simple solution - modifying the track line of the wig file 
will correct the problem, though it may not be simple for all users 
depending on what software you have available to you.  In the past, the 
solution has been as simple as moving the 'type="wiggle_0"' to follow 
immediately after word 'track'.

The problem seems to arise because FindPeaks 3.1.9.2 produces wig files 
with the "track" line parameters in a different order from that given in 
the wig file documentation, 
(http://genome.ucsc.edu/goldenPath/help/wiggle.html).  I don't know why 
only some files produce the error, through re-ordering the parameters 
does seem to resolve the issue for the two examples I've seen.

This problem has been fixed in FindPeaks 3.2, but was not back ported to 
FindPeaks 3.1. Unfortunately, FindPeaks 3.2 won't be available as a 
recompiled jar file for another week or so.

In any case, if you'd like to send me your wig file (or just the first 
10 lines, either should work, if the file isn't too big), I would be 
happy to help you find a fix for the problem.

Finally, it was also pointed out to me by my colleague that you seem to 
be using a parameter "-minimum 1".  You may wish to raise this value for 
tracks submitted to the UCSC browser, as this can cause wig files to 
become very large, and I understand the UCSC browser may not load wig 
files over 300k long. (Peak heights below 2 are rarely informative 
anyhow, and significantly add to the size of the file.)

Alternately, you might wish to leave this particular parameter off 
entirely, as a minimum peak height of 1 will still provide all peaks 
under most circumstances.

In any case, feel free to contact me if you need any further help with 
this issue.  There is also a FindPeaks mailing list at 
[EMAIL PROTECTED], which can be used for 
any other FindPeaks problems you may encounter in the future.

Cheers,

Anthony Fejes,
BC Genome Sciences Centre



Another Email contained:
> Message: 5
> Date: Mon, 27 Oct 2008 17:17:31 +0100
> From: <[EMAIL PROTECTED]>
> Subject: [Genome] ChIP-seq
> To: <[email protected]>
> Message-ID:
>     <[EMAIL PROTECTED]>
> Content-Type: text/plain;    charset="us-ascii"
> 
> Dear sir / madame,
> 
> 
> 
> I've performed Chromatin Immunoprecipitation and had a service company
> sequence the fragments with an Illumina Solexa sequencer. After the
> sequence reads had passed the quality check, the ELAND program was used
> to align the reads to the hg18 provided by UCSC as part of the Illumina
> data analysis pipeline. This was also done by the same company.
> 
> Thereafter the aligned sequence tags were fed into the FindPeaks
> 3.1.9.2. package written by the Canada Michael Smith Genome Sciences
> Centre in order to 1. analyze the short-read sequences to identify areas
> of enrichment of the protein of interest and 2. generating wig files for
> use in the UCSC genome browser. The company performed a FindPeaks run
> with standard settings. After I adjusted the settings to my own wished,
> I performed another FindPeaks run. The wig file that was generated by
> the company could be put into the genome browser without any problems.
> After uploading the wig file created by myself, it seems at first that
> it went all right because the following screen appears:
> 
> 
> 
> Manage Custom Tracks
> 
> genome: Human    assembly: Mar. 2006    [hg18]
> Name
> 
> Description
> 
> Type
> 
> Doc
> 
> test3_duplicates_subpeaks_len_triangle
> <http://genome.ucsc.edu/cgi-bin/hgCustom?hgsid=114649973&hgct_table=ct_t
> est3duplicatessubpeakslentriangle>
> test3_ht:1.0_FL:triangle_dupe_rds_inc
> 
> wiggle_0
> 
> 
> But when I want to go to the genome browser, I the following error
> message appears:
> 
> 
> 
> Internal error (Unmatched ' line 1 of input: missing closing ' ):
> removing custom tracks
> 
> 
> Do you know by any chance what went wrong? I've already tried the
> standard FindPeaks settings the company also used, but it doesn't work
> form me.
> 
> I hope someone will have an answer for me. Thanks in advance!
> 
> 
> 
> Regards,
> 
> Marije
> 
> 
> 
> Drs. Marije Telgenkamp, MD
> 
> PhD candidate Molecular Cell Biology
> 
> Department of Science and Technology
> 
> University of Twente
> 
> Building Zuidhorst, room ZH125
> 
> PO box 217
> 
> 7500 AE Enschede, The Netherlands
> 
> Phone: +31 (0)53 4893554
> 
> Fax: +31 (0)53 4891105
> 
> [EMAIL PROTECTED]
> 

_______________________________________________
Genome maillist  -  [email protected]
http://www.soe.ucsc.edu/mailman/listinfo/genome

Reply via email to