Hello, thank you for your excellent work on ComBindVS!
I’m currently studying your paper and have a question regarding the DUD-E benchmark setup used in your experiments. From the paper "Leveraging nonstructural data to predict structures and affinities of protein–ligand complexes", I understand that helper ligands were selected such that the Tanimoto similarity to the test ligand was either <0.2 or between 0.2 and 0.3. In each case, 10 helper ligands were sampled, and 39 targets in DUDE satisfied the <0.2 condition, while 76 targets satisfied the 0.2–0.3 condition.
I was wondering if you’d be willing to share — or let me know where I might find — the following information:
- The list of the 39 targets and 76 targets that satisfied each similarity criterion;
- (If available) The specific sets of 10 helper ligands sampled in each of the 5 random runs for those targets.
This information would be immensely helpful as I would like to ensure a fair comparison to ComBindVS.
Hello, thank you for your excellent work on ComBindVS!
I’m currently studying your paper and have a question regarding the DUD-E benchmark setup used in your experiments. From the paper "Leveraging nonstructural data to predict structures and affinities of protein–ligand complexes", I understand that helper ligands were selected such that the Tanimoto similarity to the test ligand was either <0.2 or between 0.2 and 0.3. In each case, 10 helper ligands were sampled, and 39 targets in DUDE satisfied the <0.2 condition, while 76 targets satisfied the 0.2–0.3 condition.
I was wondering if you’d be willing to share — or let me know where I might find — the following information:
This information would be immensely helpful as I would like to ensure a fair comparison to ComBindVS.