DrugGEN Resource Collection: Training Data, Model Weights, Generated Molecules, Docking and MD Analyses

The DrugGEN dataset repository archives the full training datasets (curated ChEMBL compounds, AKT1 and CDK2 bioactivity records, smiles correction datasets), the pre-trained weight files for all targeted and non-targeted model variants, and all result artefacts—including de novo–generated molecules, docking output tables, and Molecular Dynamics analyses.The repository is organized as follows:<b>data/</b>Raw dataset filesSMILES correction files<b>encoders/</b> — Encoder files for molecular representations<b>decoders/</b> — Decoder files for molecular representations<b>experiments/</b><b>models/</b> — Pre-trained weights (DrugGEN-AKT1, DrugGEN-CDK2, NoTarget)<b>results/</b><b>generated_molecules/</b> — De novo molecules analyzed in the paper<b>docking/</b> — Molecular docking results for de novo molecules from DrugGEN and models compared in the paper<b>MD_Data.zip</b> — MD trajectory data for selected 6 molecules (trajectories can be found at: https://drive.google.com/drive/folders/1aBRoTaacF63RjaOMihx2JpbkS3nGmvaU)<br><br>