An Evolutionary Heuristic Approach for Functional Modules Identification from Composite Biological Data

摘要

Modular structures are key features of the complex biological networks which control the functioning and dynamics of the living cell. In order to unravel this intricate modularity we propose an algorithm having evolutionary features that unveils and optimizes sub-networks within a weighted graph of protein interactions. The graph is enriched by integrating information from gene expression profiles. Functional modules are retrieved by expanding a kernel protein set that originates from 'seed' proteins and validated through biological and topological criteria. The modules obtained are further refined using a framework based on binary particle swarm optimization.