Gene Network Inference via Structural Equation Modeling in Genetical Genomics Experiments

• DOI: 10.1534/genetics.107.080069
• Publication Date: 2008-02-04T01:28:41Z
• Authors: Bing Liu; Alberto de la Fuente; Ina Hoeschele
• Keywords: Gene regulatory network

Our goal is gene network inference in genetical genomics or systems genetics experiments. For species where sequence information available, we first perform expression quantitative trait locus (eQTL) mapping by jointly utilizing cis-, cis-trans-, and trans-regulation. After using local structural models to identify regulator-target pairs for each eQTL, construct an encompassing directed (EDN) assembling all retained relationships. The EDN has nodes corresponding expressed genes eQTL edges from cis-regulated target genes, cis-trans-regulated trans-regulator trans-eQTL genes. within the strongly constrained search space defined EDN, propose equation modeling (SEM), because it can model cyclic networks indeed contains feedback On basis of a factorization likelihood space, our SEM algorithm infers involving several hundred eQTL. Structure based on penalized ratio adaptation Occam's window selection. was evaluated data simulated with nonlinear ordinary differential equations known topologies applied real yeast set.