Our lab’s investigations center around the idea that cells are computational entities containing complex biological circuitry for sensing the external environment, processing signals through networks of interacting components and producing chemical output as well as regulating state reconfiguration. We are especially interested learning predictive computational models from single-cell phosphoproteomic and transcriptomic data. To this end our previous work has focused on developing information-theoretic methods for identifying and characterizing interactive relationships between signaling proteins and uncovering how these relationships are dysregulated in disease such as cancer and diabetes.

Ongoing work involves creating more sophisticated and accurate models of transformational biological processes such as epithelial-to-mesenchymal transition in cancer, and inflammation progression in the immune system, by analyzing both signaling and genomic data. We are developing methods that learn how signaling networks dynamically rewire over extended periods of time, methods for uncovering novel causal relationships from observational and perturbation data, as well as more basic methods for probabilistically modeling single-cell transcriptomic data despite its inherent sparsity.