Designing and validating optimal nonaddictive analgesics using the CANDO paradigm

摘要 我们提出了一种新的整体范式，通过蛋白质组规模的深度学习来帮助解决阿片类药物危机 设计最佳非成瘾性镇痛药的方法。我们将设计针对特定fic蛋白的化合物。 这些途径结合了美沙酮和丁丙诺啡等长效药物的止痛效果 同时调节额外的相互作用以确保通过药物在奖励系统中的最佳药理消亡 比如纳洛酮和纳曲酮。目标蛋白质将通过对电流的多尺度分析来确定 具有额外专家信息的药物，产生所需的目标蛋白质组(即，具有 设计的候选药物应该与哪种药物相互作用以获得最佳效果和最小副作用)以缓解 痛苦而不上瘾。为达到这些目的而设计的药物将被合成并使用绑定来验证 和动物研究在第一阶段，最有希望的候选者将是继续药物开发的领导者 在第二阶段，最终目标是临床试验和监管批准。我们的范例具有广泛的适用性 设计候选药物以帮助缓解阿片类药物危机。 我们开发了新药再利用机会计算分析(CANDO)平台。 枪支多靶点药物的发现、再利用和设计，部分资金来自NIH主任的先锋奖，以 克服传统单一目标方法的局限性。开源平台的筛选和排名 通过大规模建模和分析，为每个疾病/适应症(和不良事件)提供药物/化合物 药物/化合物综合文库与蛋白质结构之间的相互作用。这些互动是 使用各种分子对接和评分方案确定。Cando实现了各种板凳- 用于猎枪重新用途的标记协议，即确定每种已知药物如何相互关联 批准的适应症/疾病的背景，从而能够对各种管道进行评估 以及平台内部和外部的协议，以用于药物发现。又快又准的多重 交互评分/对接协议、蛋白质组学规模和使用的严格的全能基准 该平台使其成为以所需蛋白质组空间为目标的化学实体设计的独特和理想选择 或客观的。具体而言，fi我们的目标是：(I)创建增强的目标互动签名/网络 描述疼痛和毒瘾。(Ii)应用深度学习管道生成药物设计并对其进行排名 所需蛋白质组学/交互作用目标的背景。(Iii)使用装订和装订验证顶部设计 动物研究。总体而言，我们的目标是对Cando中的所有药物/化合物进行分析，以设计最佳 蛋白质组学非成瘾止痛目标的合成和临床前验证。最有效的去- 我们公司将通过第二阶段的授予进一步寻求药物开发和监管批准的迹象， 风险投资和/或行业伙伴关系。

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.