us47h-LigandScoringProtocol.txt

Name

CNN docking with affinity model ranking

Software

docking performed with gnina commit b3fa6ae13fc6b42924f49b2d751d68f1bc14bc08 available from https://github.com/gnina/gnina , conformer generation performed with rdkit via https://github.com/dkoes/rdkit-scripts/rdconf.py, ensemble of receptors chosen via Pocketome.

Parameters

rdconf.py was used with the defaults except --maxconfs=1 was specified. The PDB was searched using the provided FASTA sequence, and the top resulting accession ID was used to query Pocketome to return a set of possible reference receptors for the compounds that did not bind with DMSO or SO4 in a critical bridging location. Those receptors were aligned to the provided SO4-bound structure, solvent was removed, and multimers were reduced to a single monomer for docking. Reference ligands were used to define a box, and then 8 Angstroms of padding were added to each face of that box to define the search space for docking; in the case of the DMSO and SO4 reference receptors, both of those molecules were used together for both the receptors to define a search space in the same manner. The random seed was set to 0 and the exhaustiveness was set to 50. The default number of final conformers was used, which is 9. --gpu and --cnn_scoring were enabled. Compounds were ranked by the affinity predicted by the default convolutional neural network affinity model. To be added to the final prediction set, a pose was required to differ from poses already in the set by more than 0.5 RMSD in order to ensure pose diversity.

Method

for each SMILES input, a single conformer was generated via rdkit using the UFF force field. CatS_14 was docked to the provided reference receptor containing DMSO only, and CatS_2, CatS_17, CatS_20, CatS_22, CatS_23, CatS_24 were docked to the provided reference receptor containing SO4 only; the remaining compounds were docked to each of five receptors found via Pocketome that have high sequence similarity to the provided references. These receptors were chosen due to the conformational diversity of their binding sites based on visual inspection. All compounds were docked by gnina using the default convolutional neural network scoring function.

Answer 1

No

us47h-PosePredictionProtocol.txt

Name

CNN docking with affinity model ranking

Software

docking performed with gnina commit b3fa6ae13fc6b42924f49b2d751d68f1bc14bc08 available from https://github.com/gnina/gnina , conformer generation performed with rdkit via https://github.com/dkoes/rdkit-scripts/rdconf.py, ensemble of receptors chosen via Pocketome.

System Preparation Parameters

rdconf.py was used with the defaults except --maxconfs=1 was specified. The PDB was searched using the provided FASTA sequence, and the top resulting accession ID was used to query Pocketome to return a set of possible reference receptors for the compounds that did not bind with DMSO or SO4 in a critical bridging location. Those receptors were aligned to the provided SO4-bound structure, solvent was removed, and multimers were reduced to a single monomer for docking.

System Preparation Method

for each SMILES input, a single conformer was generated via rdkit using the UFF force field. CatS_14 was docked to the provided reference receptor containing DMSO only, and CatS_2, CatS_17, CatS_20, CatS_22, CatS_23, CatS_24 were docked to the provided reference receptor containing SO4 only; the remaining compounds were docked to each of five receptors found via Pocketome that have high sequence similarity to the provided references. These receptors were chosen due to the conformational diversity of their binding sites based on visual inspection. All compounds were docked by gnina using the default convolutional neural network scoring function.

Pose Prediction Parameters

Reference ligands were used to define a box, and then 8 Angstroms of padding were added to each face of that box to define the search space for docking; in the case of the DMSO and SO4 reference receptors, both of those molecules were used together for both the receptors to define a search space in the same manner. The random seed was set to 0 and the exhaustiveness was set to 50. The default number of final conformers was used, which is 9. --gpu and --cnn_scoring were enabled. Compounds were ranked by the affinity predicted by the default convolutional neural network affinity model. To be added to the final prediction set, a pose was required to differ from poses already in the set by more than 0.5 RMSD in order to ensure pose diversity.

Pose Prediction Method

Compounds were docked using gnina's default convolutional neural network model. Docked poses were ranked by their predicted affinity. The best-scoring pose was chosen, and subsequent poses were chosen in order of rank if they differed from existing poses in the chosen set by greater than 0.5 RMSD until at most 5 poses were chosen. Compounds with unblinded crystal structures made available for this subchallenge were scored based on the crystal structures directly.

Answer 1

No

Answer 2

Yes