Designing and validating optimal nonaddictive analgesics using the CANDO paradigm

点击翻译按钮获取中文摘要

ABSTRACT We propose a novel holistic paradigm to help solve the opioid crisis by using proteome scale deep learning approaches to design optimal nonaddictive analgesics. We will design compounds that target specific proteins and pathways the combine the pain relieving effects of long acting drugs such as methadone and buprenorphine while modulating additional interactions to ensure optimal pharmacological extinction in reward systems via drugs like naloxone and naltrexone. The proteins to be targeted will be determined by multiscale analytics of current drugs with additional expert information, resulting in desired objective proteomes (i.e., the subset of proteins with which a designed drug candidate should interact with for optimal efficacy and minimal side effects) for relieving pain without addiction. The drugs designed to these objectives will be synthesized and validated using binding and animal studies in Phase I, and the most promising candidates will be leads for continued drug development in Phase II with the eventual goals of clinical trials and regulatory approval. Our paradigm is broadly applicable to design drug candidates to help alleviate the opioid crisis. We developed the Computational Analysis of Novel Drug Repurposing Opportunities (CANDO) platform for shot- gun multitarget drug discovery, repurposing, and design, funded in part by a NIH Director’s Pioneer Award to overcome the limitations of traditional single target approaches. The open source platform screens and ranks drugs/compounds for every disease/indication (and adverse event) through large scale modeling and analytics of interactions between comprehensive libraries of drugs/compounds and protein structures. The interactions are determined using a variety of molecular docking and scoring protocols. CANDO implements a variety of bench- marking protocols for shotgun repurposing, i.e., to determine how every known drug is related to every other in the context of the indications/diseases for which they are approved, which enables evaluation of various pipelines and protocols within and external to the platform for their utility in drug discovery. The multiple fast and accurate interaction scoring/docking protocols, the proteomic scale, and rigorous all-against-all benchmarking used within the platform make it unique and ideal for the design of chemical entities that target a desired proteomic space or objective. Specifically our aims are to: (i) Create enhanced objective interaction signatures/networks describing pain and addiction. (ii) Apply deep learning pipelines to generate and rank drug designs in the context of desired proteomic/interactomic objectives. (iii) Validate the top designs using binding and animal studies. Overall, our goal is to perform analytics of all drugs/compounds within CANDO to design optimal proteomic nonaddictive analgesic objectives that are synthesized and preclinically validated. The most potent de- signs will be pursued further by our company for drug development and regulatory approval via a Phase II award, venture capital and/or industry partnerships.