Computational Aspects of Random Boolean Networks

Speaker: Elena Dubrova

Abstract:

Research on networks becomes essential to all branches of sciences as we struggle to interpret the data coming from neurobiology, genomics, economics, ecology, and the Internet. Random Boolean Networks (RBNs) were introduced by Kaufmann in 1969 in the context of gene expression and fitness landscapes. They were applied to the problems of cell differentiation, immune response, evolution, and neural networks. They have also attracted the interest of physicists due to their analogy with the disordered systems studied in statistical mechanics, such as the mean field spin glass. An RBN is a synchronous Boolean automaton. Each vertex has k predecessors, selected at random, and an associated Boolean function of k variables. Kauffman has shown that it is possible to tune the parameters of an RBN so that the network exhibits self-organized critical behavior ensuring both stability and evolutionary improvements. Statistical features of self-organized RBNs match the characteristics of living cells. This talk focuses on computational aspects of RBNs. First, we give an introduction to random Boolean networks and show how they can be used for modeling of gene regulatory networks of living cells. Then, we describe three basic steps of the analysis of dynamical behavior of RBNs: redundancy removal, partitioning, and computation of attractors. Finally, we discuss open problems and outline prospectives of RBNs.