How to calculate methylation FC for genes based on CpG FC data?

Question

[maria2019]

Hi,

I have done methylation analysis on a treatment vs. control and have a list of CpGs, FC, FDR, gene names, ..etc.

I am interested in comparing the methylation based on genes. Since there are sometimes multiple CpGs in genomic regions, is there any way that I can report a methylation FC for genes?

For example, if I have 2 CpGs with FC of 3 and 2 at gene A, can I say that methylation FC for gene A is (3+2)/2=2.5?

Answer 1

[Devon Ryan]

You can certainly take the mean or median, though a fold change for methylation data makes no sense. What you care about is the actual change in methylation, which can never be greater than 1. A 3 fold change in methylation is biologically meaningless if it's from 0.01 to 0.03, just as an example.