Siddiqui Benjamin, Yadav Chandra Shekhar, Akil Mohd, Faiyyaz Mohd, Khan Abdul Rahman, Ahmad Naseem, Hassan Firoj, Azad Mohd Irfan, Owais Mohammad, Nasibullah Malik, Azad Iqbal
Department of Chemistry, Integral University, Lucknow, India.
Department of Laboratory Animal Facility, CSIR-CDRI, Lucknow, India.
Comb Chem High Throughput Screen. 2025 Jan 15. doi: 10.2174/0113862073334062241015043343.
Computer-Aided Drug Design (CADD) entails designing molecules that could potentially interact with a specific biomolecular target and promising their potential binding. The stereo- arrangement and stereo-selectivity of small molecules (SMs)--based chemotherapeutic agents significantly influence their therapeutic potential and enhance their therapeutic advantages. CADD has been a well-established field for decades, but recent years have observed a significant shift toward acceptance of computational approaches in both academia and the pharmaceutical industry. Recently, artificial intelligence (AI), bioinformatics, and data science have played a significant role in drug discovery to accelerate the development of effective treatments, reduce expenses, and eliminate the need for animal testing. This shift can be attributed to the availability of extensive data on molecular properties, binding to therapeutic targets, and their 3D structures. Increasing interest from legislators, pharmaceutical companies, and academic and industrial scientists is evidence that AI is reshaping the drug discovery industry. To achieve success in drug discovery, it is necessary to optimize pharmacodynamic, pharmacokinetic, and clinical outcome-related properties. Moreover, the advent of on-demand virtual libraries containing billions of drug-like SMs, coupled with abundant computing capacities, has further facilitated this transition. To fully capitalize on these resources, rapid computational methods are needed for effective ligand screening. This includes structure-based virtual screening (SBVS) of vast chemical spaces, aided by fast iterative screening approaches. At the same time, advances in deep learning (DL) predictions of ligand properties and target activities have become very helpful, as they no longer need information about the structure of the receptor. This study examines recent progress in the drug discovery and development (DDD) approach, their potential to reshape the entire DDD process, and the challenges they face. This review examines the role of artificial intelligence as a fundamental component in drug discovery, particularly focusing on small molecules. It also discusses how AI-driven approaches can expedite the identification of diverse, potent, target-specific, and drug-like ligands for protein targets. This advancement has the potential to make drug discovery more efficient and cost-effective, ultimately facilitating the development of safer and more effective therapeutics.
计算机辅助药物设计（CADD）需要设计出可能与特定生物分子靶点相互作用并有望产生潜在结合作用的分子。基于小分子（SMs）的化疗药物的立体排列和立体选择性显著影响其治疗潜力并增强其治疗优势。几十年来，CADD一直是一个成熟的领域，但近年来，学术界和制药行业都出现了向接受计算方法的重大转变。最近，人工智能（AI）、生物信息学和数据科学在药物发现中发挥了重要作用，以加速有效治疗方法的开发、降低成本并消除动物试验的需求。这种转变可归因于有关分子特性、与治疗靶点的结合及其三维结构的大量数据的可用性。立法者、制药公司以及学术和工业科学家日益增长的兴趣证明，人工智能正在重塑药物发现行业。为了在药物发现中取得成功，有必要优化药效学、药代动力学和临床结果相关特性。此外，包含数十亿个类药物小分子的按需虚拟库的出现，再加上丰富的计算能力，进一步推动了这一转变。为了充分利用这些资源，需要快速的计算方法来进行有效的配体筛选。这包括借助快速迭代筛选方法对广阔化学空间进行基于结构的虚拟筛选（SBVS）。与此同时，深度学习（DL）对配体特性和靶点活性的预测进展非常有帮助，因为它们不再需要受体结构的信息。本研究考察了药物发现与开发（DDD）方法的最新进展、它们重塑整个DDD过程的潜力以及所面临的挑战。这篇综述考察了人工智能作为药物发现基本组成部分的作用，尤其关注小分子。它还讨论了人工智能驱动的方法如何能够加速为蛋白质靶点鉴定多样、强效、靶点特异性和类药物的配体。这一进展有可能使药物发现更高效、更具成本效益，最终促进更安全、更有效的治疗方法的开发。
