I've characterized the number of peaks binding to each genomic location. In my Chipseq experiment I've counted the number of peaks that bind to intergenic, promoter-transcription start sites, introns, exons, transcription termination sequences, untranslated regions etc. Is there a way to tell if there is enriched binding to these regions?
Right now im looking at a CEAS: cis-regulatory element annotation system as a way to compare chipseq data to background data. However I need to look in how feasible this is for a matched chipseq experiment
This thread hits on two separate issues. First, how do know which ChIP-Seq peaks are real binding sites, as opposed to sequencing artifacts (i.e. ones that would show enrichment for IgG)? Second, is your genomic mark associated with a specific genomic feature, such as proximal promoters, transcription start sites, distal enhancers, etc?
For the first question, run the IgG as well as input (pre-ChIP) controls. For the second, CEAS does a nice job. If you're familiar with R, ChIPpeakAnno, which is a BioConductor package, is another good option.
Some useful guidelines for beginners here: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3828144/
Nice illustration (common mistakes deliberately included) of the how to get started with R and BioConductor, here: http://moo.nac.uci.edu/~hjm/R_BioC_example.html
Hi steph,
if you are"nt familiar with R, there are a galaxy application of CEAS, in the CISTROME galaxy application,
u can also find other analysis
http://cistrome.org/ap/root
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools