Infer Gene Regulatory Networks from Time Series Data with Probabilistic Model Checking

Gene regulatory relationships constitute a complex mechanism of interactions adopted by cells to control behaviours and functions of a living organism. The identification of such relationships from genomics data through a computational approach is a challenging task as the large number of possible solutions is typically high in contrast to the number of available independent data points. Literature approaches address the problem by reducing the search space and/or extend the amount of independent information. In this paper we propose a probabilistic variant of a previous proposed approach based on formal methods. The method starts with a formal specification of gene regulatory hypotheses and then determines which is the probability that such hypotheses are explained by the available time series data. Both direction and sign (inhibition/activation) of regulations can be detected whereas most of literature methods are limited just to undirected and/or unsigned relationships. We empirically evaluated the probabilistic variant on experimental and synthetic datasets showing that the levels of accuracy are in most cases higher than those obtained with the previous method, outperforming, indeed, the current state of art.