Integrative analysis of methylation and gene expression data in TCGA

Abstract—An integrative analysis of methylation and gene expression is performed using TCGA data for 997 samples from multiple cancer types. We use conditional mutual information to examine each gene separately, identify 798 genes whose expression are repressed by their methylation, and derive gene-specific thresholds to dichotomize their methylation data. For each methylation controlled gene, we evaluate its conditional correlation with other genes, and infer which other genes are its regulators or targets. The resulting set of regulator genes is much larger than that of target genes, and the set of regulator genes is highly enriched by transcription factors. I