metadata
dataset_info:
features:
- name: ID
dtype: string
- name: Sequence
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 3650734.2822767235
num_examples: 9724
- name: test
num_bytes: 405845.71772327623
num_examples: 1081
download_size: 3975390
dataset_size: 4056580
Dataset Card for "Diffusion-Evodiff_knots"
Dataset of proteins from Evodiff
Dataset of proteins generated through Evodiff
Contains knotted proteins (tested with omegafold and topoly) and unknotted proteins in ratio 1:4
The dataset is splitted into a train set (majority, 90 %) and evaluation set (only 10 %)