Donwload our source data into this directory in this repository. There are two files in this directory: download.sh and untar.sh.

• download.sh is a script for downloading all the datasets.
• untar.sh is a script for extracting all the downloaded files.

By executing download.sh, you can download all the following datasets in tarfiles directory.

training dataset

• PDBbind v2020 scoring: preprocessed pdbbind v2020 refined set
• PDBbind v2020 docking: preprocessed re-docking augmented dataset
• PDBbind v2020 cross: preprocessed cross-docking augmented dataset
• PDBbind v2020 random: preprocessed random-docking augmented dataset

benchmark dataset

• CASF-2016 socring: preprocessed CASF-2016 scoring benchmark dataset
• CASF-2016 docking: preprocessed CASF-2016 docking benchmark dataset
• CASF-2016 screening: preprocessed CASF-2016 screening benchmark dataset
• DUD-E screening: preprocessed DUD-E screening benchmark dataset (docked with smina, single conformation)
• derivative benchmark: preprocessed derivative benchmark 2015 dataset

After that, you can extract the downloaded files by executing untar.sh. This will make two new dataset, PDBbind-v2020 and Benchmark, in this directory.

Also, you can generate PDA data with codes in generate_PDA