lhallee/Full_PDB_Contacts
收藏资源简介:
--- 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.
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.
数据集概述
数据集特征
- id: 数据类型为字符串。
- seqs: 数据类型为字符串,具有序列属性。
- contacts: 数据类型为无符号16位整数,具有序列属性。
数据集分割
- train: 包含90,193个样本,总大小为3,896,443,787字节。
数据集大小
- 下载大小: 1,539,844,300字节。
- 数据集大小: 3,896,443,787字节。
配置
- config_name: default
- data_files:
- split: train
- path: data/train-*




