Chip seq annotation

[cambridgetim]

Hi All. This is my first post on seq answers.

I am trying to figure out how to annotate a bed file containing peaks, telling me which peaks are located next to which genes.

I have figured out three options by Googling:

1. Cistrome peak2gene option

2. Bioconductor's ChIPpeakAnno

3. Also, this is a little more tenuous, but I gather I can UCSC's table browser, make a bed file containing locations of genes and use bedtools to compare this with the bed file containing my peak data.


Any advice on how these 3 approaches compare?

Link to table browser here:

http://genome.ucsc.edu/cgi-bin/hgTables?hgsid=216014043&clade=worm&org=0&db=0&hgta_group=genes&hgta_track=knownGene&hgta_table=knownGene&hgta_regionType=genome&position=&hgta_outputType=primaryTable&hgta_outFileName=

[ETHANol]

This is a really nice work flow.

http://fasta.bioch.virginia.edu/cshl/stubbs/galaxy_bed2genes.html

Doesn't exactly annotate your bed files but it does give you a list of genes of which your protein of interest binds to, which I am guessing is what you want.

[cambridgetim]

Thanks ETHANol. I need a little more context. Or more specifically, my boss is going to ask for details like:

1. how does this method compare to other methods?
2. how do most people normally do this?

Thanks for your help.

[ETHANol]

1. This isn't one of those things like mapping or peak calling where there are qualitative/qualitative differences between methods. This methods simply give you the promoters of genes your txt factor binds to. There may be other methods that save you time.

2. I have no idea how 'most people' do it. I'm sure a lot of people do it differently. The key is finding a workflow that works for you and your data set. For instance you mention a Bioconductor package. I find everything from Bioconductor to be a real pain to figure out. Other people are really good with R-based workflows so they'll go straight to them.

Figure out the question you want to answer and devise a work flow that brings you to the answer.