Boris
Open Babel 2.4.1/chimera alpha version 1.14/Pafnucy version 1.0
Grid resolution of 1A
Standard approach, By training a Convolutional Layer on ~10k compounds binding affinities, I obtained a model that I twisted the architecture somewhat to obtain a new model (different from the original one)
I used it to rank the ligands For this one, I used the protein structure as a whole instead of extracting the AS (defined as 8A around the ligand --> 2HHN PDB files)
No
Yes