From signaling cartoons to model-based inference in cancer immunology

04/22/2011 14:00
America/New York
Speaker: David J. Klinke II, West Virginia University

Venue: Mount Sinai School of Medicine, Annenberg Conference Room 20-01

Abstract:
Cells sense and respond to their external environment through a network of protein-protein interactions. This flow of information within cells is summarized by a cell signaling network. These signaling networks ubiquitously appear in the literature as cartoons that enumerate the key proteins and their interactions that control this flow of information. Implicitly a signaling cartoon provides a framework, i.e., a prior, for interpreting biological data. However, it is unclear how well a specific cartoon applies to a specific cell type or to a disease state, like cancer. Techniques for quantitatively evaluating this prior information in light of new data are just beginning to emerge. The iterative process of updating our knowledge, given our prior knowledge and the specific data at hand, corresponds to reasoning in a Bayesian framework. These techniques come at a critical time as our uncertainty in how specific cells process information is a barrier for the rational design of drugs that manipulate cell fate. In this talk, I will summarize some of our recent efforts in understanding how cells make decisions from data using Bayesian model-based inference. Specific questions are related to cancer immunology and include re-wiring of signaling networks in cancer cells, signaling control mechanisms in T helper cells, and mechanisms of immunoescape.