DOI: 10.5176/2251-2489_BioTech15.39
Authors: Mrityunjay Sarkar and Aurpan Majumder
Abstract: Identification of genes in differential modules having unique gene interaction patterns across different conditions is an important problem drawing a lot of attention in recent years. Genes having differential connectivity may play a pivotal role in complex diseases. So identification of these differentially wired genes could lead to a better and suitable drug target against those diseases. Traditional methods consider either only the differential connectivity or differential expression to identify significant genes, whereas in this work we have proposed an extended version of an existing concept called topological overlap (TO) by not only considering the change in the connectivity but the changes in the expression levels too. Here, we have implemented the generalized topological overlap measure (GTOM) to find out a neighborhood of genes similar to a seed gene across different conditions. Based on the GTOM result we have gone through weighted TO analysis which shows inter-dependency of all genes present in a network as well as un-weighted TO showing dependency of only some significantly connected genes in a network. Finally, we fetched those genes having low TO value as well as low p-value (from permutation test/t-test) and checked their significance. Accordingly, investigating the depth of the problem in hand we could explore the fruitfulness of the proposed approach.
Keywords: Peripheral blood; Influenza; Topological Overlap (TO); Generalized Topological Overlap Measure (GTOM); G.O; KEGG pathway; weighted TO; un-weighted TO
Updating...