The FEM R package: Identification of Functional Epigenetic Modules

This vignette provides examples of how to use the package FEM to identify interactome hotspots of differential promoter methylation and differential expression, where an inverse association between promoter methylation and gene expression is assumed [1]. By “interactome hotspot” we mean a connected subnetwork of the protein interaction network (PIN) with an exceptionally large average edge-weight density in relation to the rest of the network. The weight edges are constructed from the statistics of association of DNA methylation and gene expression with the phenotype of interest. Thus, the FEM algorithm can be viewed as a functional supervised algorithm, which uses a network of relations between genes (in our case a PPI network), to identify subnetworks where a significant number of genes are associated with a phenotype of interest (POI). We call these “hotspots” also Functional Epigenetic Modules (FEMs). Current functionality of FEM works for Illumina Infinium 450k data, however, the structure is modular allowing easy application or generalization to DNA methylation data generated with other technologies. The FEM algorithm on Illumina 27k data was first presented in [2], with its extension to Illumina 450k data described in [1]. The module detection algorithm used is the spinglass algorithm of [3]. The PIN used in this vignette includes only proteinprotein interactions, derives from Pathway Commons [4] and is available from http://sourceforge.net/projects/funepimod/files under filename hprdAsigH*.Rd, but the user is allowed to specify his own network. There are three main components to this vignette. These are: