This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Help with HOMER motif profile plots

Hi,

I'm trying to understand profile plots for motifs. Homer has a function annotatePeaks.pl) to generate motif density profiles that looks like following (lifted from here)

enter image description here

While I understand what these profile plots convey for ChIP-seq data (e.g. those generated by deepTools plotProfile), but I'm not sure what exactly is the y-axis in above plot.

So the question is, is it just the probability of finding given motif at each position (as MEME does here) in a set of given sequences or is it the actual count normalized to something?

I didn't find any explanation on homer website. It would be great if anyone can explain what exactly is going on in above plot.

Thank you.

homer motifs homer
• 4,638 views • link
written by venu venu

Convert a BED file of motif positions to a bigWig with a 1 where there's a motif and a 0 otherwise and it's the same thing you'd then get with plotProfile I think.

• 1 views • link
written by Devon Ryan Devon Ryan

I created a bigiwg ( using bedtools genomecov + kent utils) and plotted with plotProfile. The motif I'm interested in is has very sparse hits, applying loess fit on plotProfile output data is more/less looking closer to the above plot. I think I understood what's happening inside.

• 1 views • link
written by venu venu

1 answer

For future googlers, here is the response from lead author of HOMER, Chris Benner

The motif plots are normalized by taking the actual motif totals, then normalizing by the bin size used in the histogram (so that the motif densities are comparable across different bin sizes) and the total number of peaks/regions used to create the histogram (so that the motif densities are independent of the number of regions used to search).

• 1 views • link
written by venu venu

Log in to answer this question.

More posts like this