Bayesian Learning of Genetic Network Structure in the Space of Differential Equation Models

D. Peavoy,A. Kabán

Abstract:Introduction. Reverse engineering of genetic regulatory networks (GRNs) from the expression levels of thousands of genes over time is a central problem in systems biology. A number of approaches have been developed, with effort from the biology, statistics and machine learning communities. The difficulty lies with the very large number of variables and unknowns involved. Gene interactions exhibit complex networks of intra-cellular regulation and inter cellular communication, with highly non-linear dynamics. All modelling approaches necessarily approximate the true complexity to some extent, in order to proceed with limited computational resources. However, to gain understanding of the mechanisms of a system (rather than aiming at black-box prediction only), a more detailed and more realistic description of the true dynamics is required.