Modeling biological networks serves as both a major goal and an effectivetool of systems biology in studying mechanisms that orchestrate the activitiesof gene products in cells. Biological networks are context specific and dynamicin nature. To systematically characterize the selectively activated regulatorycomponents and mechanisms, the modeling tools must be able to effectivelydistinguish significant rewiring from random background fluctuations. Weformulated the inference of differential dependency networks that incorporatesboth conditional data and prior knowledge as a convex optimization problem, anddeveloped an efficient learning algorithm to jointly infer the conservedbiological network and the significant rewiring across different conditions. Weused a novel sampling scheme to estimate the expected error rate due to randomknowledge and based on which, developed a strategy that fully exploits thebenefit of this data-knowledge integrated approach. We demonstrated andvalidated the principle and performance of our method using synthetic datasets.We then applied our method to yeast cell line and breast cancer microarray dataand obtained biologically plausible results.
展开▼
