This proposal aims to complement LINCS data and other publically available datasets by generating the molecular profile data necessary to transform combination therapy from a trial and error exercise to a rational discipline, by allowing identification of synergistic drug combinations on a predictive basis. Specifically, we will establish a platform for the rational, quantitative study of synergistic drug activity not just for cell viability but for an entire range of disease-relevant cellular phenotypes. This pilot study will (a) generate a set of experimental assays to characterize phenotypic transitions in a given cell type, (b) prioritize synergistic drug combinations using a systems biology approach, based on existing expression and 1050 profiles from LINCS and Connectivity Map (CMap) data, (c) test combinations predicted to be synergistic versus random combinations to assemble an objective benchmark to estimate performance of predictive algorithms, and (d) characterize the synergistic behavior of drugs with orthogonal Mechanism of Action (MoA) vs. drugs sharing a similar MoA, by pairing a small panel of drugs with well-established MoA, with compounds that have a predicted orthogonal MoAs across different cell lines. By orthogonal, we mean MoAs that share no substrates and effectors. While the main aim of this proposal is the generation ofthe experimental data necessary to harness drug synergy, its value also lies in its integration of computational Systems Biology methods for the assessment of synergistic activity of drug combinations toward the implementation of several functional cellular phenotypes, rather than just viability. This is predicated on an extensive framework of systems biology tools for the assembly and interrogation of context-specific molecular interaction maps (interactomes) developed in the Pi's lab. It should also be noted that, except for of a handful of kinase inhibitor combinations, no multiperturbation assays are being tested in LINCS centers. Thus, this is an important unmet need ofthe LINCS program.
While the discovery of new therapeutics is of paramount importance in the fight against diseases such as cancer, significant improvements in patient outcome could be achieved if drugs in the current FDA-approved repertoire could be optimally matched to individual malignancies on a predictive basis. The goal of this proposal is to leverage data available through the LINCS program in order to predict and validate optimal and synergistic drug combinations to achieve a given phenotypic outcome.
| Alvarez, Mariano J; Shen, Yao; Giorgi, Federico M et al. (2016) Functional characterization of somatic mutations in cancer using network-based inference of protein activity. Nat Genet 48:838-47 |
| Woo, Jung Hoon; Shimoni, Yishai; Yang, Wan Seok et al. (2015) Elucidating Compound Mechanism of Action by Network Perturbation Analysis. Cell 162:441-451 |
| Chiu, Hua-Sheng; Llobet-Navas, David; Yang, Xuerui et al. (2015) Cupid: simultaneous reconstruction of microRNA-target and ceRNA networks. Genome Res 25:257-67 |
| Bansal, Mukesh; Yang, Jichen; Karan, Charles et al. (2014) A community computational challenge to predict the activity of pairs of compounds. Nat Biotechnol 32:1213-22 |
