International Journal of Scientific Research and Engineering Development

AI is modifying how we make new drugs and improve healthcare. The traditional process of developing new drugs takes years and involves a few billion dollars, but AI allows the process to be more rapid, accurate, and affordable. In this paper, we summarize five papers published between 2020-2024 to see which AI technologies have provided the most interesting results in drug development. We review methods, including, but not limited to, machine learning for in silico screening, deep learning-based natural product identification, generative AI to create new drugs structures, AI to decode patient data, and transformer-based deep neural networks. We discuss transformer-based models, such as BioBERT, ChemBERTa, and AlphaFold, which perform acceptably well in predicting molecular properties and predicting drug-target interactions in cancer studies and protein structure prediction. Using this AI as a foundation, we present a method that generates an AI model that uses transformer-based molecular structure modelling, in conjunction with patient data profiling using electronic health records. This approach could aid in the rapid development of new drug candidates, reverse the clinical trial failure rate, and assist physicians in providing personalized, effective care to their patients.