Novel combination therapies are potentially key to preventing emergence of resistance to initially successful single agent therapies such as RAF inhibitors in melanoma. In order to systematically nominate novel and effective drug combinations, we have developed an integrated systems biology technology (perturbation biology) for inferring quantitative and predictive network models of signaling. Our modeling strategy combines systematic perturbation experiments, measurement of response profiles to perturbations, inference of network models and simulation of models with in silico perturbations. As a key component of the overall strategy, we have adapted the belief propagation (BP) inference algorithm from statistical physics to construct signaling models with several hundreds of measured entities. In parallel, we have developed a pathway informatics tool to automatically extract prior information from multiple signaling databases. We applied this approach to derive signaling network models in RAF inhibitor resistant melanoma cells and nominate drug combinations to overcome resistance in melanoma. First, we systematically perturb melanoma cells with a set of targeted drugs as single and paired agents. Next, we quantitatively measured proteomic (total and phospho-protein levels) and phenotypic (e.g. cell cycle arrest, cellular viability) responses to perturbations. We incorporated the generic prior information and context specific perturbation response data to construct quantitative network models, which capture multiple oncogenic pathways in melanoma. As shown by cross validation calculations, use of prior information significantly improved predictive power of models. Using the inferred network models of signaling, we systematically predicted response to tens of thousands of in silico perturbations. Our modeling and simulation strategy expanded the volume of drug response profile from few thousand experimental data points to millions of in silico data points. In addition, our models provided quantitative descriptions of signaling events in melanoma. Our perturbation biology technology is suitable for applications in diverse areas of molecular biology beyond cancer research. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):A24. Citation Format: Anil Korkut, Weiqing Wang, Evan Molinelli, Martin Miller, Poorvi Kaushik, Arman Aksoy, Xiaohong Jing, Nick Gauthier, Chris Sander. Network models of signaling and drug response in melanoma. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr A24.
Use of targeted drug combinations are potentially key to preventing emergence of resistance to initially successful single agents such as RAF inhibitors. Thus, systematic nomination of novel drug combinations is important to develop effective therapeutic strategies. Here, we present an experimental-computational network pharmacology method to infer quantitative signaling networks in tumor cells, predict response to perturbations, and ultimately, nominate targeted drug combinations that will generate a desired phenotypic response. We use a series of targeted drugs, singly or in combination to perturb cancer cells. We measure quantitative proteomic (total and phoshpo-protein levels) and phenotypic (e.g. cell viability, apoptosis, cell cycle progression) response profiles to perturbations. Proteomic responses are measured with the reverse phase protein array technology. Next, we infer quantitative network models, using the response profiles as training sets. Solution space of all network model configurations is prohibitively large and Monte Carlo based inference algorithms fail to generate accurate network models in sizes relevant for cancer biology. In order to solve the inference problem, we have adapted an iterative and probabilistic inference algorithm, belief propagation from statistical physics. Resulting network models are based on simple non-linear differential equations and quantitatively link signaling events to phenotypic changes. We have generated quantitative network models of signaling in RAF inhibitor resistant melanoma cell lines in sizes (i.e. ∼100 nodes) unreachable by other network inference algorithms. We have nominated novel combination therapies through combinatorial in silico perturbations of all nodes in derived networks and experimentally tested the predicted phenotypic responses. Citation Format: Anil Korkut, Weiqing Wang, Evan Molinelli, Martin Miller, Poorvi Kaushik, Arman Aksoy, Chris Sander. Quantitative network models of signaling and drug response in melanoma. [abstract]. In: Proceedings of the 104th Annual Meeting of the American Association for Cancer Research; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2013;73(8 Suppl):Abstract nr 5227. doi:10.1158/1538-7445.AM2013-5227
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.