Machine learning-driven drug repurposing study to identify new tubulin inhibitors against cancer

This dataset contains all machine learning models, molecular dynamics simulation trajectories, and docking poses generated during our study on AI-driven drug repurposing for the identification of novel colchicine-binding site (CBS) inhibitors of tubulin with anticancer potential. The files include AutoQSAR model outputs, docking poses of the 3 best inhibitors (omeprazole, podofilox and sulfadoxine) as well as their 200 ns MD simulation trajectories, along with the MD files of the reference co-crystallized inhibitor G8K. These resources are provided to support reproducibility, facilitate further analysis, and support future research on computational drug discovery targeting tubulin.