Methylation statistical analysis

Question

[dBlandford]

So I've been given a JMP file containing methylation percentage data for a few subjects. The file is organized by columns, containing the chromosome number, position on chromosome, and percentage of methylation for that location for each subject (15 subjects total).

chr    pos     subj
chr1   10485   9.2208

I tried going over a few bioconductor packages for R as well as different software for methylation analysis, but found no one that can handle this kind of data. The purpose is to run an analysis (similar to a gwsa using Plink, for exmaple) in order to find differences in methylations between the study groups.

Any advice on either how to organize the file to a more proper file format or which software will be appropriate for the situation?

Thanks

[Devon Ryan]

Was the source of this data a methylation array or something else? You'll almost always have better luck with the raw data rather than something that someone exported from JMP.

Answer 1

[Sean Davis]

You could take a look at the limma package, as it handles differential methylation (similar to gene expression). Alternatively, take a look at Bumphunter.

As for data format, you can load the data into R and simply manipulate it there into the data structure as needed.

[Devon Ryan]

Note that data should be logit transformed for use in limma (or pretty much anything else).