In this application we propose to build a new in silico platform (dubbed MetaDrugTM)for predicting metabolism and possible toxic effects of novel drug candidates. The platform will integrate advanced QSAR and expert system approaches with our extensive database on human pathways and our software for reconstruction and visualization of metabolic and cell signaling networks. First we develop capabilities to analyze compound's structural and physicochemical similarity to known substrates of all human Cytochrome P450 superfamily enzymes and predict whether and how it may enter human metabolism. Using our existing reaction database we will identify specific types of biochemical transformations catalyzed by human cytochromes (model reactions) and formulate sets of most essential molecular features and rules for every such type. We will develop a special algorithm capable to recognize such rules and apply them (in conjunction with advanced QSAR methods) to identify the likely metabolites of any novel xenobiotic compound. Second we will adapt our existing network-building software for predicting the metabolic fates of xenobiotics. We will develop a novel tool that generates sets of putative reactions based on predicted substrate potential and model reactions. The existing network building software will handle these newly introduced reactions and visualize them as interactive maps. Third we develop an interface that aligns predicted networks with our functional maps of human biochemistry and cell signaling. The user will be able to view interactions between xeno- and endobiotics metabolism pathways in different tissues, analyze them visually on functional maps, access related information on associated diseases, genetic polymorphisms, etc. When completed, this platform will be applicable in predicting the toxicity of novel drug candidates, evaluating the risks associated with environmental pollutants and monitoring human exposure to pathogens and toxins.
| Embrechts, Mark J; Ekins, Sean (2007) Classification of metabolites with kernel-partial least squares (K-PLS). Drug Metab Dispos 35:325-7 |
| Ekins, Sean; Kirillov, Eugene; Rakhmatulin, Eugene A et al. (2005) A novel method for visualizing nuclear hormone receptor networks relevant to drug metabolism. Drug Metab Dispos 33:474-81 |
| Ekins, Sean; Andreyev, Sergey; Ryabov, Andy et al. (2005) Computational prediction of human drug metabolism. Expert Opin Drug Metab Toxicol 1:303-24 |
| Balakin, Konstantin V; Ekins, Sean; Bugrim, Andrey et al. (2004) Kohonen maps for prediction of binding to human cytochrome P450 3A4. Drug Metab Dispos 32:1183-9 |
| Balakin, Konstantin V; Ekins, Sean; Bugrim, Andrey et al. (2004) Quantitative structure-metabolism relationship modeling of metabolic N-dealkylation reaction rates. Drug Metab Dispos 32:1111-20 |
