metadata
dataset_info:
features:
- name: pdb_id
dtype: string
- name: SeqA
dtype: string
- name: SeqB
dtype: string
- name: kd
dtype: float64
- name: pkd
dtype: float64
- name: labels
dtype: float64
splits:
- name: train
num_bytes: 15623
num_examples: 31
download_size: 20061
dataset_size: 15623
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Processed version of the haddock protein-protein affinity benchmark from here.
We only include the sequences and affinity if an example is involed with only two chains in a pdb file.
Please cite their work if you use this dataset. There is parsed information about the sources here.