The inference of molecular networks from “omics” data is one of the central problems in systems biology, the so-called reverse-engineering problem. Particular emphasis is on the reverse- engineering of gene regulator y networks from DNA microarray data. There is a growing body of literature focused on the development of efficient algorithms for this purpose. Important features of such algorithms should be robustness to data noise and the ability to incorporate prior biological knowledge. REACT (Reverse Engineering Algorithm with Evolutionary Computation Tools) is a software application to reverse engineer Boolean networks. It takes as input one or more time courses of microarray data and produces an optimal Boolean network model of the underlying gene regulator y network. The algorithm is able to incorporate both wildtype and perturbation data, as well as prior biological knowledge. In addition to a wiring diagram for the network, the algorithm also provides qualitative information about the network dynamics. The method consists of an evolutionary algorithm (EA) that optimizes the dynamic network structure with respect to data fit and a novel numerical measure of model complexity.
 
Paola Vera-Licona’s Homepage