Coarse-grain reconstruction of genetic networks from expression levels

L. DIAMBRA

PHYSICA A - STATISTICAL AND THEORETICAL PHYSICS

ELSEVIER SCIENCE BV

Año: 2011 vol. 390 p. 2198 - 2207

0378-4371

In the postgenome era many efforts have been dedicated to systematically elucidate thecomplex web of interacting genes and proteins. These efforts include experimental andcomputational methods. Microarray technology offers an opportunity for monitoring geneexpression level at the genome scale. By recourse to information theory, this study proposesa mathematical approach to reconstruct gene regulatory networks at a coarse-grainlevel from high throughput gene expression data. The method provides the a posterioriprobability that a given gene regulates positively, negatively or does not regulate each oneof the network genes. This approach also allows the introduction of prior knowledge andthe quantification of the information gain from experimental data used in the inferenceprocedure. This information gain can be used to choose those genes that will be perturbedin subsequent experiments in order to refine our knowledge about the architecture of anunderlying gene regulatory network. The performance of the proposed approach has beenstudied by in numero experiments. Our results suggest that the approach is suitable forfocusing on size-limited problems, such as recovering a small subnetwork of interest byperforming perturbation over selected genes.