CrossDocked2020
A commonly used benchmark dataset in structure-based drug design and molecular generation, constructed from the CrossDocked dataset through cleaning and re-docking. It contains on the order of one hundred thousand protein-ligand complex structures, with training and test sets split by protein sequence similarity to evaluate model generalization to unseen pockets.
Timeline (last 90 days)
CrossDocked2020 is a commonly used benchmark dataset in the SBDD and molecular generation fields, containing on the order of one hundred thousand protein-ligand complex structures, and is split into training/test sets by protein sequence similarity to ensure that the test pockets differ sufficiently from the training pockets at the sequence level.
This study uses one complete run log on pockets from the CrossDocked2020 test set to examine whether the keys of the experience library recur on unseen pockets and still distinguish improving edits from worsening edits.
All Facts (2)
CrossDocked2020 is a commonly used benchmark dataset in the SBDD and molecular generation fields, containing on the order of one hundred thousand protein-ligand complex structures, and is split into training/test sets by protein sequence similarity to ensure that the test pockets differ sufficiently from the training pockets at the sequence level.
50%UnverifiedThis study uses one complete run log on pockets from the CrossDocked2020 test set to examine whether the keys of the experience library recur on unseen pockets and still distinguish improving edits from worsening edits.
50%