Learning signaling pathway structure from high throughput, multivariate single cell data

Learning signaling pathway structure from high throughput, multivariate single cell data

Karen Sachs

Stanford University

Cells respond to their environment via signaling pathways, in which extracellular cues trigger a cascade of information flow, causing signaling molecules to become chemically, physically or locationally modified, gain new functional capabilities, and affect subsequent molecules in the cascade, culminating in a phenotypic cellular response. The disregulation of signaling pathways can be found across the entire spectrum of human disease, most notably in cancer. We have previously demonstrated the elucidation of signaling pathway structure using high dimensional flow cytometry, a technology that enables measurement of multiple signaling pathway molecules in thousands of single cells in high throughput (Sachs et al, Science, 2005). From thousands of single cell measurements taken under various stimulus conditions, we extracted the signaling pathway structure using a probabilistic approach called Bayesian networks. This tool elucidates multivariate probabilistic dependencies from large scale datasets, which may represent underlying mechanistic regulatory relationships. We present here extensions to Bayesian networks that we have developed to overcome limitations in this approach and to expand our ability to analyze signaling pathways, in healthy and disease states.