SAMPL4 Binding Blind Challenge

Our group achieved top marks on the SAMPL4 HIV integrase blind challenge. The results will made public soon in a special volume of the Journal of Computer Aided Molecular Design.

The integrase enzyme of the HIV virus is responsible for incorporating the viral genome into the chromosome of the infected cell. In SAMPL4, participants from all over the world used computational models to predict binders of integrase from a focused pool of candidate ligands. Prediction performance was judged based on experimental results which were not known prior to submission of the predictions. Our method involved docking and large scale binding free energy calculations with the BEDAM method and it achieved the best enrichment factor among all computational submissions.

Below is a video of the "alchemical" BEDAM molecular dynamics trajectory of one of the integrase complexes. In this segment the ligand complex undergoes a unbinding-binding-unbinding transition. Note how the protein and the ligand "freeze-up" when they interact. The concomitant entropic loss plays an important role in regulating binding.

Our group has also received top marks for the prediction of binding free energies of supramolecular complexes, like the one with a octa-acid molecular container shown here. Binding free energy calculations in this SAMPL4 category have been performed by students using the BEDAM method