First, the user enters up to five compounds to the Virus-CKB platform. A task is assigned for each compound and the input format of the compounds are converted to PDB (Protein Data Bank) and PDBQT (similar to PDB) formats.

Next, the Blood-Brain Barrier predictor, HTDocking and the TargetHunter are applied to the compounds simultaneously. The BBB predictor predicts the ability of compounds to penetrate the BBB. Docking scores are generated by putting the compounds into previously defined binding pockets in “different conformations of a target protein.” The TargetHunter calculates the similarity score of molecules against the known active compound dataset.

Then, using docking scores and the similarity score, a target classification is performed and these results are passed through Spider Plot for visualization.