Machine Learning in Drug Discovery and Development
作者
Narendra Pal Singh,Abhay Gangawr
标识
DOI:10.1109/smart66937.2025.11389598
摘要
This is more important that how to reduce the human workload and enhance quality of life significantly as number of diseases are increase day to day. For this purpose, the current rise of machine learning in drug discovery and development has been playing a significant role. This article basically describes two main tasks in drug discovery and development. First one is target identification and hit generation to absorption, distribution, metabolism, excretion and toxicity (ADMET) prediction and clinical trial optimization. Second one is how ML models especially deep learning, graph neural networks and generative models are increasing timelines, improving hit quality and enabling data-driven decisions. This paper categorizes various machine learning methodologies, common datasets and benchmarks, representative applications such as virtual screening, de-novo design, retrosynthesis, ADMET and industrial successes. The discussion also covers specific persistent challenges such as data quality, interpretability, domain shift and regulatory concerns in drug discovery and development. The findings cover in this paper should help researchers, students, and the pharmaceutical industry learn more about machine learning and artificial intelligence in the context of drug discovery and development. They also show how machine learning techniques are widely used in pharmaceutical research, drug discovery and development and also point to a bright future for these technologies.