A0790
Title: Integrative structural learning of mixed graphical models via pseudo-likelihood
Authors: Qingyang Liu - University of Connecticut (United States)
Yuping Zhang - University of Connecticut (United States) [presenting]
Abstract: Markov Random Field is a common tool to characterize interactions among a fixed collection of variables. In recent biomedical research, new concerns have arisen about the discovery of regulatory and co-expression relationships among different types of features across multiple biological classes. Consequently, a data integration framework is proposed to learn multiple mixed graphical models simultaneously and jointly. To address the common asymmetry problem in neighbourhood selection, a new estimator is constructed using regularized pseudo-likelihood, which produces sym- metric and consistent estimates of network topologies. The practical merits of the method are demonstrated through learning synthetic networks and constructing gene regulatory networks from TCGA data.
