Inferring Causal Gene Regulatory Networks Using Time-Delay Association Rules

Abstract

Inferring cause and effect relationship in Gene Regulatory Networks (GRNs) is a vital and a challenging topic. In this paper, we generate association rules between a pair of genes to reconstruct GRN from expression data. While computing confidence of a rule, we emphasize on the fact of time delay involved during the process of regulation between a target factor gene and any target gene. We, generate strong binary rules to infer GRN from expression data. We use in-silico DREAM challenge data for experimentation and assessments, due to the availability of ground truth networks. We compare our results with few well known causality inference methods and outcomes are satisfactory. Our results confirm the fact that inference of causality in GRN is a time delay activity and should be taken care during inference process.