Advancements in drug discovery: integrating CADD tools and drug repurposing for PD-1/PD-L1 axis inhibition.

Despite significant strides in improving cancer survival rates, the global cancer burden remains substantial, with an anticipated rise in new cases. Immune checkpoints, key regulators of immune responses, play a crucial role in cancer evasion mechanisms. The discovery of immune checkpoint inhibitors (ICIs) targeting PD-1/PD-L1 has revolutionized cancer treatment, with monoclonal antibodies (mAbs) becoming widely prescribed. However, challenges with current mAb ICIs, such as limited oral bioavailability, adverse effects, and high costs, underscore the need to explore alternative small-molecule inhibitors. In this work, we aimed to identify new potential ICI among all FDA-approved drugs. We employed QSAR models to predict PD-1/PD-L1 inhibition, utilizing a diverse dataset of 29 197 molecules sourced from ChEMBL, PubChem, and recent literature. Machine learning techniques, including Random Forest, Support Vector Machine, and Convolutional Neural Network, were employed for benchmarking to assess model performance. Additionally, we undertook a drug repurposing strategy, leveraging the best