compare the background levels in two chip-seq experiments

Question

[Bogdan]

Dear all,

which algorithms would you recommend in order to assess the background levels between two ChIP-seq experiments ?

Thanks,

Bogdan

Answer 1

[seidel]

It might depend a little on what you consider background, this could refer to background signal in the IP sample, or background peaks found in the input (whole cell extract) sample. However, there are a few common ways to assess differences in samples.

1. You could examine the Fraction of Reads in Peaks (Landt et al, Genome Research Sept. 2012, 22(9): 1813–1831).
2. You could examine the Irreproducible Discovery Rate (also mentioned in Landt et al.)
3. Depending on how you found your peaks, if you used MACS for instance, you could examine the distribution of P or Q-Values of the peaks. Different levels of background would give rise to different distributions for your peaks, as well as comparing the fold enrichment values for each peak set.