Entering edit mode
8.3 years ago
morovatunc
▴
550
Dear all hi,
From my chip-seq data I got bed and bw format of same data.
Say:
When I looked these files with IGV, the peak heights are not matching. Not exactly same.
Is this before, during bw to BED transformation, they calculated the are under BW and directly recoded as a single hight of the whole peak instead continues values to reduce storage?
I explained the difference based on this transformation. Was I wrong? Could someone explain me what actually happens during the format transformation?
You'd need to look at the bigWig and BED files to find this out, no one that doesn't have the files can answer this for you.
Which peak caller are you using?
I did not call peak by myself. However, they used mac14
The BED file reports the binding regions predicted by MACS, and are usually displayed in a browser as a horizontal thick black line. In MACS14 the BIGWIG (BW) file is generated from a WIG file that contains the coverage of DNA fragments every other 10bp of the genome, by default. This can be viewed as a profile of peaks and troughs. In MACS2 coverage plots are generated as BEDGRAPH files where coverage is reported for a range of bases with the same value, rather than every base, or 10bp. Sorry for the delay, just say your response.
Usually this is what happens: you call peaks and get a bed file, you can convert this to a bedgraph/wig file, this one is converted into a big wig.
The bed file contains information about the peak location plus some metrics specific to the peak caller like fold change and p-value. In the bigwig file only the location and one of those values is kept.
When you load this data into IGV, the bed files looks like horizontal bars spanning the width of the peak. In contrast, the bigwig file has a height that is proportional to the measure it contains (for example p-value, or fold change). But you will have to ask to know which value is that.