There is increasing interest in understanding the pathological role of DNA methylation changes in disease by profiling genome-wide methylation changes. This includes both 5-methylcytosine (5mC) and 5-hydroxymethylcytosine (5hmC). The typical profiling study is designed to measure 5mC and/or 5hmC levels alongside gene expression in a set of samples and controls to determine a list of candidate genes whose 5mC and/or 5hmC changes are associated with expression changes. We recently showed that ME-Class2 substantially outperforms other bioinformatic approaches at accurately identify genes with highly associated methylation and expression changes. ME-Class2 further illuminated how synergistic changes in 5mC and 5hmC potentially contribute to gene silencing and activation. Here we present a detailed protocol for using ME-Class2 to analyze genome-wide methylation (5mC and/or 5hmC) and expression data. Further, we provide advice about extending ME-Class2 to study the relationships between other epigenetic marks.