A cloud-based approach for Gene Regulatory Networks dynamics simulations

Gene Regulatory Networks (GRNs) are one of the most investigated biological networks in Systems Biology because their work involves all living activities in the cell. A powerful but simple model of such GRNs are Boolean Networks (BN) that describe interactions among biological compounds in a qualitative manner. One of the most interesting outcomes about GRNs's dynamics are the so called network attractors, since they seem to well represent the stable states of a living cell. Though collecting state space trajectories is a quite simple task when the network topology consists of few nodes, it becomes not so trivial when nodes are of the size of hundreds or thousands. Thus, we exploit the MapReduce algorithm in order to cope this complexity on a cloud architecture built for the purpose. We found that scaling-out the problem is a better solution rather than increasing resources on single machine, thus allowing simulations of large networks.