I have some data I am working with, and I am curious if I am able to combine p-values from a paired t-test for CpG sites in the genome using Fisher's Method to get one p-value for each unique gene. Linked here is the Wikipedia page for Fisher's Method. I understand that an assumption of the method used is that each individual p-value being combined must be independent. I'm relatively new to biostatistics, so I'm curious if using CpGs from the same gene would violate this assumption.

  • $\begingroup$ Yes, it would very likely to violate the assumption for any two pvalues computed on data coming from the same chromosome. You might want to read this section of the wiki article to understand how violation of these assumptions may affect your results. However, one would need to know how you computed your pvalues in the first place to be definitive. How did you compute your pvalues? $\endgroup$ – Remi.b Aug 10 '17 at 15:52
  • $\begingroup$ p-values were computed using a paired t-test before and after treatment $\endgroup$ – user35655 Aug 10 '17 at 15:57
  • $\begingroup$ You have to explain a bit more in order to understand whether these pvalues may be independent. What is the response and explanatory variable (or the two groups in your t-test if you prefer) and what is your treatment. But you probably have a good intuition of the biology and of the treatment performed yourself without having to explain of whether your data from one site to the next are independent or not. $\endgroup$ – Remi.b Aug 10 '17 at 16:12
  • 3
    $\begingroup$ This question may be better received in Bioinformatics.SE. Yes? No? $\endgroup$ – user22020 Aug 10 '17 at 16:21
  • $\begingroup$ Ok I will elaborate as much as possible without divulging too much for the sake of confidentiality. The explanatory variable is a drug treatment given to patients. There are many potential response variables, but the one I am interested is an M-value, which is a log2 ratio of methylated to unmethylated probes detecting a single CpG site. Does that make sense? $\endgroup$ – user35655 Aug 10 '17 at 16:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.