A Comparative Approach for Multiple Gene Network Inference Using Time-Series Gene Expression Data
Monday, October 20, 2003 - 3:00pm - 3:20pm
Guillaume Bourque (University of Montreal)
We present a method for gene network inference and revision based on time-series data. Gene networks are modeled using linear differential equations and a generalized stepwise multiple linear regression algorithm is used to recover the interaction coefficients. Our system was design for the recovery of gene interactions concurrently in many gene regulatory networks related by a graph or a tree. Suppose we are studying a certain regulatory network in different species of known phylogeny. We can think of the different networks as being related to each other in that way and use this information. Alternatively, we might be interested in the development stages of this network or we could be studying the same system but in different tissues related at a different level. The idea is that, given gene expression data for each species, or each stage of development, or each tissue, we seek to recover each individual network while minimizing a cost based on the differences along the edges of the graph or the tree. We show how this comparative framework allows new insights and facilitates the gene network inference process.