Phenotypic constraints drive the architecture of biological networks

by Areejit Samal

Friday 14 Jun 2013, 14:30 → 16:00 Europe/Rome

Aula Conversi (Dip. di Fisica - Edificio G. Marconi)

Biological networks have architectural features that distinguish them from random networks such as degree distribution, high clustering, high robustness and over-representation of certain network motifs. In this talk, I ask whether such unusual features follow from phenotypic constraints on specific networks. The standard benchmark ensemble or null model used to detect unusual features in biological networks is the edge-randomization algorithm. I will present our method based on Markov Chain Monte Carlo (MCMC) sampling to generate realistic benchmark ensembles for metabolic networks and gene regulatory networks wit! h a given phenotype. I then redefine the notion of unusual features of biological networks using this realistic benchmark ensemble. I consider E. coli metabolic network and Boolean gene regulatory model for Arabidopsis flowering as two example systems from different levels of biological organization. By generating realistic benchmark ensembles for the two systems, I show that most unusual structural properties of these networks could arise due to functional constraints. I conclude that function is a main driver of biological network structure.