Gonad Methylation Analysis Part 7

M&M: methylKit and MACAU

Now that I’ve run samples through bismark, my next step is to analyze the methylation data produced. This can be done with methylKit or MACAU in R. Steven tasked with installing both programs and reading the manuals, so I did just that (kind of).

M1: methylKit methylKit is made for reduced-representation bisulfite sequencing, but can apparently handle whole-genome bisulfite sequencing data (i.e. my data) if it’s in the “right format.” I installed the package on the Mac mini in this R script. It was however, extremely confusing and difficult because data.tables (a dependency) wouldn’t properly compile. I think I was able to figure it out by just entering “no” every time it asked me to do something, but I’ll have to revisit it.

The manual is detailed, but confusing. I don’t think I’ll get a feel for the order of operations until I actually start using it. I could definitely refer to Steven’s lab notebook post to figure it out!


MACAU somehow makes more sense to me, even though it’s all about modeling (thanks, Skalski). It fits a binomial mixed model to the data to identify differential methylation. However, it can only run on a linux machine. I’ll probably run methylKit first on the Mac mini, then jump over to MACAU.

One thing I need to figure out before starting is what the difference between these two programs are. I’ll probably also look at the methods sections of methylation papers just to get a clearer understanding of what I’m doing.

Written on May 1, 2018