Measure genome-wide agreement of peaks in
Is there an approach to measure agreement in epigenomic signals e.g. histone marks (continuous -log10 p-vals after peak calling)? This approach should also work across assays or cell types. For example, how similar is the DNAse-seq signal across Neutrophils, Monocytes and T-cells? Are these more similar than the DNase-seq from a Neuron? Essentially I’m looking for a measure of overlap of peaks on a genome-wide scale (I have heard of people selecting different peaks/regions of the genome to determine this). I believe simple correlation be largely effected by sparsity of the data. Does other factors need to be taken into account like sequencing depth? Any pointers to literature of approaches on this would be massively helpful!
Note I know I could find the proporion of overlapping peaks but this doesn’t take into account the continuous score assigned to each peak.
• 13 views
Read more here: Source link