In an epigenome-wide association study, a large number (100+) of significant differentially methylation positions (DMPs), would be identified, in contrast to GWAS, where most signals would be confined to small number of loci.
As a bioinformatician, when assessing the significant signals, I would just query the loci one by one manually to search for potential biological processes behind the trait under investigation since there would be just a few of them.
However, in EWAS, this is certainly not feasible as there could be more than 100 signals. I tried looking into the genes the DMPs mapped to according to the manifest file (by Illumina), but it is unclear how the CpG sites were mapped to that gene, and how methylation will affect the functions of the target gene. I tried pathway or GO term enrichment, but what I got were vague terms and I don't know how to follow up on enrichment analyses. I also tried looking for DMPs that are eQTMs, but eQTM resources are hard to find, and many DMPs are not eQTMs.
What are your general strategies in assessing DMPs to 1) find out DMPs that are worth pursuing and 2) search for the biological processes behind the DMPs?
