Entering edit mode
7.3 years ago
igor
13k
I noticed that sometimes WGBS/RRBS experiments have what I call "incomplete" methylation (there is probably a better term for it).
Here is an example (generated by MethylKit):
Notice how the unmethylated side of the graph is okay, but the methylated side drops off at the end. The last bin should be highest, but it's not. The rest of the metrics seem okay as far as I can tell. Is that a problem? Is there an explanation?
Well one could argue that given a large population of cells, one would expect a highly methylated site to still be randomly unmethlyated a small percentage of the time just due to random biological noise.
I considered that as well, but shouldn't that affect both ends of the spectrum? Since there are more methylated CpGs, it should affect the unmethylated end of the spectrum even more.
Which kind of sample it is? it is so weird, never seen such kind of distribution before. RRBS or WGBS?
I've seen this with both WGBS and RRBS.