Probabilistic refinement of cellular pathway models

1. Probabilistic models called Nested Effects Models can reconstruct signaling pathways from high-dimensional phenotypic profiles obtained after perturbing candidate pathway genes. 2. These models account for the "information gap" between directly observing effects on pathway components versus indirectly observing downstream effects on reporter genes. 3. The models place candidate pathway genes in a graph representation that best explains the observed phenotypic profiles, refining our understanding of information flow within signaling pathways.