TAME-VS: Target Expansion and Virtual Screening

Retrieve compounds related to the expanded targets from ChemBL and collect SMILES structures.

Convert compound SMILES into molecular fingerprints (Morgan/ECFP) and generate a feature matrix for machine learning.

Train machine learning models with molecular fingerprints to build compound activity predictors.

Use trained models to run virtual screening on compound libraries and predict candidate activity.

Post-process virtual screening results, including binding mode analysis and ADMET prediction.

Integrate module outputs and generate the final design report and downloadable result package.