Hello! Recently I sent some samples to a company for ChIP-seq against H3K79me2 and I want to compare the differentially methylated regions between four samples, two of them treated with the DOT1L inhibitor SGC0946. On the average plot, the two treated samples (named with XX-SGC) clearly have lower methylation, however, the differential analysis results using the merged regions show that the drug-treated samples have more hypermethylated regions compared to the untreated samples (as shown on the bar plot). Just wondering why I am seeing those hypermethylated regions and shouldn’t there be more hypomethylated regions in the drug- treated samples? I am sure the plot was made correctly. Thank you for answering!
If the inhibitor globally (in genome-wide) decreases/increases histone methylation level, you will need a spike-in control (usually use fly chromatin). The methylation intensity signals/counts should normalize to the spike-in before any downstream analyses including signal profiling and differential analysis.
I have some introduction here:
https://spiker.readthedocs.io/en/latest/introduction.html
But you can easily find more papers on this topic.