chipseq data analysis

Question

[praveenhcu131]

how to calculate the h3k4me3 breadth in chipseq data it is oe of the feature which i want to study please tell me how to calculate the peak breadth

Answer 1

[Devon Ryan]

You have two options:

1. Call peaks and determine the widths from the BED file.
2. Forget peaks and plot the distribution around the TSS. You'll see a large peak there, so just measure the width.

[Deepak Tanwar]

Which peak caller would you suggest @Devon?

[Petr Ponomarenko]

MACS2 is very popular in general. BCP has better sensitivity. MUSIC is more sensitive for broad peaks like h3k4me3.

[praveenhcu131]

can you please tell me how to plot the distribution around tss from macs out put files

[Devon Ryan]

You wouldn't, you'd plot the read density (make a bigWig file with bamCoverage) and plot that (computeMatrix and then plotHeatmap or plotProfile).

[praveenhcu131]

sir can u tell detailed procedure how to do that

[Devon Ryan]

I already mentioned the exact tools (they're all part of deepTools). You can even do that in Galaxy if you prefer. The deepTools documentation is fairly decent.

[praveenhcu131]

i have done plotpca and plot profile from now where i have to measure the breadth of peak

[Devon Ryan]

The normal options would be:

• Eye ball it
• Use photoshop to measure the width
• Use a ruler

[praveenhcu131]

i have to measure the width from transcription start site to end site how can i measure it from png image

[Devon Ryan]

If your peak is over the whole gene body then you can't get that information from plotProfile, you'll need to get it from your BED/GTF file. I presumed that you had some enrichment in the promoter or around the TSS, whose width you could just approximate by eye.