Annotation of Reads...

[Giles]

Can anyone out there help me find a program or software suite that will take the reads of a ChIP seq or an RNA seq and classify the region of the genome for each read? For example, the final output would be a pie chart depicting what percent of the reads are from rDNA, tRNA genes, miRNA genes, exons, introns, repeats (and specifically, which type of repeats, etc.). I see this type of data presented all the time, but can not figure out how to generate that once a ChIP seq (or RNA seq) is completed.

[steven]

Are your reads mapped? If so you can use Galaxy (online) or BEDtools (locally) to compare them with annotated genomic regions. Does that make sense?

[Giles]

thanks

Yes, my reads are aligned. They are in a Wiggle format. I will check those two sites out. Thanks.

[yjhua2110]

you can try UCSC Table Browser, http://genome.ucsc.edu/cgi-bin/hgTables

[houhuabin]

Are you looking for this?
http://seqanswers.com/forums/showthr...light=mirtools]

[steven]

Wiggle format? mhm.. sounds like a data point format (coverage depth for each genomic position or so). I guess that this would not let you easily get the *number* of reads in each region. You can still get a coverage value for each region from such data, but that is not the same though.
What do the others think?

Anyway, try uploading your data in Galaxy. You will probably need the annotation files of the tRNA, rRNA, miRNA etc genes, along with the positions of introns, coding exons, etc. This can be retrieved from the UCSC tables browser (use the "send to Galaxy" export option).
Once your reads and the annotation data sets are all in Galaxy, you can obtain the distribution of a) either individual reads: number of reads that overlap each category ("operate on genomic intervals" may be what you need). or b) average coverage within each region ("get genomic scores - aggregate datapoints").
Does that help?
Someone has an alternative solution?

[colindaven]

Steven, I agree. Galaxy is a good option to do this.

The best starting point is the SAM format output from your aligner. I don't think you can convert back to this from Wiggle so you might need to get the original alignments.

Upload the annotation to Galaxy, convert to interval format.
Do the same for your reads.

Then do some calculations :
I find the "Join" function particularly useful for reads per gene.

[steven]

Thanks Colin.
Giles, just one thing: keep in mind that the genomic sets of annotated positions (tRNA, miRNAs, introns, UTR exons, CDS exons, Alu repeats, etc) are usually not exclusive: since there are Alu in exons, miRNAs in UTRs, exons in introns (different transcripts), etc.. you may want to subtract the sets from each other first. Otherwise you may end up with a pie chart like this one (courtesy of faux news)

[Giles]

Thanks Everyone, I'm off and running. Galaxy is really great for those of us who can comprehend what a program does (PERL script, for example) but suck at actually writing a code.