metadata
dataset_info:
features:
- name: id
dtype: string
- name: seqs
sequence: string
- name: contacts
sequence:
sequence: uint16
splits:
- name: train
num_bytes: 3896443787
num_examples: 90193
download_size: 1539844300
dataset_size: 3896443787
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
All PDB proteins as of 5/24/24. Contacts recorded in "COO" format, where a contact gets a tuple (i,j) if the ith and jith amino are within 8 angstrom. In huggingface datasets format are a nonredundant selection of 90 thousand. The entire 200+ thousand are uplaoded as pickle files.