What can overlapping peaks in ChIP-seq samples tell?

Question

[bwd]

Consider a big population of ChIP-seq experiments (not necessarily replicates). For instance, all belonging to same tissue under similar condition (e.g., all are ALL) but belonging to different people. Lets suppose all these samples are sequenced similarly, and peaks are called all using one software with same parameter setup.

• How would you interpret overlapping peaks?

• Some sites may have few overlapping peaks, and some sites may have at least one peak from all the samples; how do you read this?

• Any research article explaining such (at least vaguely similar experiment)?

Answer 1

[goodez]

I would say it depends on the type of experiment. Most importantly what antibody are you using to detect regions of chromatin interacting with a certain protein. This will have a big impact on the variation, broadness, and abundance of peaks.

I would expect mostly overlapping peaks between individuals receiving identical treatments. Differences between the samples would be largely due to differential pathways of regulation (e.g. one sample exhibits over-expression of stress response genes while another sample does not). If all the samples are truly identical in their conditions, then they should all share the same profile of peaks due to the high level of conservation within a single species.