This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Chipseq Data And Ncrna Sequence

I have a non coding rna sequence. I also have chiq-seq data for a Transcription factor. I know I can use bowtie to map the chip-seq reads to the sequence I'm interested in. However, is there program that can quantify the coverage of each nucleotide (number of reads that cover a nucleotide) in the ncRNA sequence based on the TF chip-seq data? What Im trying to do is get the signal of TF binding by averaging the coverage of all nucleotides in my ncRna sequence.

Thanks!

deleted-post

Do you have the the genomic coordinates of your ncRNAs?

yes chr12: 54356096 - 54362515

So if you have these coords, why not just run your "normal" ChIP-seq pipeline (a peak caller) then hunt for a peak in your region(s) of interest?

as an aside, chip-seq protocol typically includes a step where you treat with RNAse to degrade all RNA so if you the TF of interest is binding on your ncRNA you might want to look into CLIP-seq instead for RNA-binding protein (http://en.wikipedia.org/wiki/CLIP-Seq)

1 answer

You might try bedtools. It has a method called coverageBed for examining coverage across segments of interest, and it can take BAM files (aligned reads) as input. So one workflow would be to map your reads with bowtie, convert the SAM output to a BAM file, then put your ncRNA coordinates in a BED file and use coverageBed to query the coverage:

coverageBed -abam aligned_reads.bam -b ncRNA.bed

However, this is a rather odd way to evaluate ChIP Seq data. How will you know if the coverage is significant versus other regions of the genome? How will you correct for background? I suppose you could put several different kinds of regions in your bed file, and perform the analysis with an input DNA as well as an IP sample, and that might tell you something along with the total number of mapped reads, but this isn't a very complete way of evaluating the data.

Log in to answer this question.