Questions About Running Differential Methylation Analysis on Paired Samples

[DKaukonen]

Hi,

First off, I would like to thank you for developing SeSAMe. I did DNA methylation years ago and found it difficult. Your package makes it so much easier and intuitive.

One thing that is not clear in the tutorials and documentation is how to handle match/paired data. We have 2 different paired analysis we are doing with over 800 people. This includes cases (cancer) and control (no known cancer), with 2 sample times for most cases.

The first is aged matched cases and controls, where we have a control with the same age as a case (for the second sampling time) for each case. Is there a way to provide DML with the list of the matched pairs? Or would adjusting for age be sufficient enough with the model that is called?

The second is similar, but not easily adjusted for. In the patients (case only) where we have 2 samples from each patient (~2 years apart) we would like to compare the first sample (before diagnosis) with the second sample (after/at diagnosis) to see which sites are more likely to be deferentially methylated. While this could be done one at a time for each patient, correction for multiple testing would be an issue. Is there a way to utilize DML on all sample pairs to identify deferentially methlyated regions between the first and second time point? In other words, how easy is it to use SeSAMe to do differential methylation analysis on paired before and after samples?

Thank you,

• Damien