Toward the accurate prediction of adverse drug reactions and drug-drug interactions using novel MM methods and QM-derived rules

药物不良反应（adr）和毒性是药物发现和开发项目中观察到的高损耗率的主要原因。研究表明，细胞色素p450 （CYPs）等代谢酶产生的反应性代谢物与蛋白质、DNA或谷胱甘肽等生物分子进一步发生反应，导致肝毒性或DNA突变导致癌症。与此同时，药物-药物相互作用（DDI）通常是由其中一种共同给药药物抑制CYP引起的。反应性代谢物和ddi的计算预测是更成功的药物发现计划的有希望的途径。在此背景下，我们开发了一个程序，impact，预测药物的代谢位点和药物与CYPs的结合。impact结合了基于预先计算的药物位点活化能的配体反应性规则和使用分子力学（MM）程序的对接。我们建议开展额外的质量管理调查，并开发改进的分子力学方法，作为我们之前与阿斯利康（2010-2012，开发影响）和化学计算小组（2013-至今）成功的CRD资助申请的延续。在过去的3年中，我们完成了原理验证研究，并将进一步开发我们的软件，包括ddi预测模块和反应性代谢物预测的改进版本，以及基于化学原理的广义和更准确的力场（FF）。

Adverse drug reactions (ADRs) and toxicity are major causes of the high attrition rates observed in drug discovery and development programs. It has been shown that reactive metabolites produced by metabolic enzymes such as cytochrome P450s (CYPs) further react with biomolecules such as proteins, DNA or glutathione, leading to hepatotoxicity or DNA mutations causing cancer. In parallel, drug-drug interaction (DDI) is often caused by CYP inhibition by one of the co-administered drugs. Computational prediction of reactive metabolites and DDIs is a promising avenue for more successful drug discovery programs. In this context, we have developed a program, IMPACTS, which predicts the site of metabolism of drugs and the binding of drugs to CYPs. IMPACTS combines ligand reactivity rules based on pre-computed activation energies of drug sites and docking using molecular mechanics (MM) routines. We propose to carry out additional QM investigations and to develop an improved molecular mechanics approach as a continuation of our previously successful CRD grant applications with AstraZeneca (2010-2012, development of IMPACTS) then chemical computing group (2013-now). Over the last 3 years we completed proof-of-principle studies and will further develop our software to include a module for DDIs prediction and an improved version for reactive metabolites prediction, as well as a generalized and more accurate force field (FF) based on chemical principles.