Integrating Artificial Intelligence Across the Drug-Discovery Pipeline: From Targets to Preclinical Proof

Authors : Dilhan Namlı, Nesrin Gökhan Kelekçi

Pages : 821-866

Doi:10.55262/fabadeczacilik.1786185

View : 68 | Download : 186

Publication Date : 2025-10-30

Article Type : Review Paper

Abstract :Artificial intelligence (AI)-based approaches have drawn significant attention for their potential to address major challenges in drug discovery processes, such as time constraints, high costs, and low success rates. Specifically, machine learning (ML) and deep learning (DL) algorithms are effectively utilized in various stages of drug development, including target identification, molecular screening, lead compound selection, optimization, and ADMET prediction. In this study, the integration of current AI models into pharmaceutical R&D processes is examined from an interdisciplinary perspective, and their application domains are evaluated through relevant case studies in the literature. It has been observed that ML-based methods can yield successful results even with limited data, while DL architectures offer advantages in modeling complex molecular relationships. Furthermore, the architectural frameworks, training strategies, and diversity of ML and DL algorithms are comprehensively discussed within the scope of the study. It is demonstrated that traditionally experience-based decision processes such as retrosynthetic planning and formulation development can be accelerated and made more sustainable through data-driven systems. Additionally, AI-assisted predictions are shown to reduce the experimental burden and enhance research efficiency in preclinical and clinical stages. These evaluations suggest that AI technologies are not merely supportive tools but also strategic components at the core of innovative drug discovery approaches.
Keywords : Yapay Zekâ, İlaç Tasarımı, Öngörüsel Modelleme, Hesaplamalı İlaç Keşfi, Sanal Tarama

ORIGINAL ARTICLE URL