The explosion of large-scale, high-throughput technologies in the biological sciences demands a comprehensive view of biological systems.   Dr. Zhu’s research focuses on:

1. Method development in integrating diverse high-throughput data into probabilistic causal network models. There are multiple risk factors as well as their interactions contributing to complex human diseases, such genetic background, infection, environmental states, life-style choices, and etc.   It is critical to integrate all these factors into comprehensive disease networks which provide a context to understand mechanism of diseases and to study effects of genetic perturbations, drug perturbations, and their interactions.  Dr. Zhu’s group has developed many methods for integrating genetic and genomic data into Bayesian networks:

• Zhu et al.,  PLoS Bio.  2012, Stitching together multiple data dimensions reveals interacting metabolomic and transcriptomic networks that modulate cell regulation.
• Tu et al., Genome Res. 2009, Integrating siRNA and protein-protein interaction data to identify an expanded insulin signaling network.
• Zhu et al., Nat. Genet. 2008, Integrating large-scale functional genomic data to dissect the complexity of yeast regulatory networks.

2. Method development in comparing multiple networks. There are multiple risk/perturbation factors driving a biological system from a healthy state to a disease state.  In response, there are hundreds to thousands genes changed at transcription or protein levels.  Dr. Zhu’s works showed that instead of monitoring changes in a few genes, it is more robust to monitor changes of network states or structures when studying complex diseases:

• Narayanan et al., PLoS Comput Biol. 2010, Simultaneous clustering of multiple gene expression and physical interaction datasets.
• Zhu et al.,PLoS Comput Biol. 2010,  Characterizing dynamic changes in the human blood transcriptional network.
• Wang et al., PLoS Comput Biol. 2009,  Meta-analysis of inter-species liver co-expression networks elucidates traits associated with common human diseases.

3. Method development in analyzing biological networks. If comprehensive biological networks are available, how to convert network knowledge into actionable items?  For example in drug discovery and development, we investigated which nodes and node combinations in the networks can generate largest beneficial effects when perturbed.  In biomarker research, we examined which parts of networks can robustly connect to drug response, disease progression and survivals.

• Wang et al, Mol Syst Biol. 2012, Systems analysis of eleven rodent disease models reveals an inflammatome signature and key drivers.
• Leonardson et al,  Hum Mol Genet. 2010,The effect of food intake on gene expression in human peripheral blood.
• Chen et al., Nature. 2008, Variations in DNA elucidate molecular networks that cause disease .