Methylation levels at specific sites

Question

[firatuyulur]

Hi all, I am working on a project where I need to check H3K79 methylation levels at enhancer sites. I have the enhancer locations as a bed file, I have the wiggle and bed file of methylation ChIP-Seq. Is there a way I can generate a count data out of wiggle format? I am thinking of processing the raw data from the start and get the bam files,with the enhancer bed file I already had, use bedtools to count the reads, then do DESeq2 for differential methylation levels.

If there is an alternative way to do this from wiggle and bed files(both from ENCODE Project) I would save some time from rerunning the alignment for 15 samples.

GSM605318 GSM605319 GSM605320 and goes on like this.

[mforde84]

This may help, not sure: https://www.r-bloggers.com/using-encode-methylation-data-rrbs-in-r/

[firatuyulur]

it is annoying that the bed files used as example has 4 or 5 extra columns than what my bed file has. therefore, the suggested package could not help me.

Answer 1

[Alex Reynolds]

If you want to count overlaps of methylation regions over enhancer regions, you could use BEDOPS wig2bed and bedmap --count:

$ wig2bed < methylation.wig > methylation.bed
$ bedmap --echo --count --delim '\t' enhancers.bed methylation.bed > answer.bed

Or, avoiding the time cost of creating an intermediate BED file:

$ bedmap --echo --count --delim '\t' enhancers.bed <(wig2bed < methylation.wig) > answer.bed

If there is other signal in the fifth column of methylation.bed, then you could replace --count with --sum, --mean, etc. to calculate density, mean, etc. of methylation signal over enhancer regions. See bedmap --help for options.