A new network-based method for gene enrichment analysis: detection of new biological processes associated to OMIM diseases

Enrichment analysis is a widely applied tool for determining functional associations between biological processes or pathways and genes or proteins related to the same phenotype. This procedure is useful for shedding light on the molecular mechanisms and functions at the basis of the analysed phenotype, for enlarging the dataset of possibly related genes and proteins and for helping interpretation and prioritisation of new experimentally determined variations. Standard enrichment methods routinely rely on the only annotations that characterize the genes/proteins included in the input. Rarely they take into consideration the physical and functional relationships among different genes or proteins that can be extracted from the available biological networks of interactions. Here we describe a method that combines the standard enrichment technique with a new procedure based on the analysis of the sub-networks connecting genes or proteins that share the same functional annotation. These sub-networks, derived from STRING, are generated by finding the minimal graphs that connect the all the proteins that share the same Gene Ontology annotation (Biological Process). We test the ability of the network-based enrichment method in finding annotation terms disregarded by a standard enrichment method and analyse 244 sets of proteins associated to different OMIM diseases. In 149 cases (61%) the network-based procedure extracts GO terms neglected by the standard method and in 79 cases (32%) some of the enriched GO terms are not included in the annotations of the original protein set.