graphs-datasets/PROTEINS
收藏资源简介:
`PROTEINS`数据集是一个中等规模的分子属性预测数据集,主要用于预测分子是否为酶的二元分类任务。数据集包含1113个图,平均每个图有39.06个节点和72.82条边。每个图的数据字段包括节点特征、边索引、边属性和标签。数据集的结构和属性信息可以通过PyGeometric库加载和使用。
The `PROTEINS` dataset is a medium-scale molecular property prediction dataset, primarily used for binary classification tasks whose objective is to predict whether a molecule is an enzyme. The dataset contains 1113 graph structures, with an average of 39.06 nodes and 72.82 edges per graph. Each graph's data fields include node features, edge indices, edge attributes, and labels. The structural and attribute information of the dataset can be loaded and utilized via the PyGeometric library.
数据集概述:PROTEINS
数据集描述
数据集总结
- 类型: 中等分子属性预测数据集
- 用途: 预测分子是否为酶,属于二分类任务
- 评估指标: 准确率,使用10折交叉验证
支持的任务和排行榜
- 任务: 分子属性预测
- 排行榜: Papers with code leaderboard
数据集结构
数据属性
- 规模: 中等
- 图数量: 1113
- 平均节点数: 39.06
- 平均边数: 72.82
数据字段
node_feat(列表: #nodes x #node-features): 节点特征edge_index(列表: 2 x #edges): 边索引edge_attr(列表: #edges x #edge-features): 边特征y(列表: 1 x #labels): 标签num_nodes(整数): 节点数量
数据分割
- 来源: PyGeometric版本,由TUDataset提供
附加信息
许可信息
- 许可状态: 未知
引用信息
@article{10.1093/bioinformatics/bti1007, author = {Borgwardt, Karsten M. and Ong, Cheng Soon and Schönauer, Stefan and Vishwanathan, S. V. N. and Smola, Alex J. and Kriegel, Hans-Peter}, title = "{Protein function prediction via graph kernels}", journal = {Bioinformatics}, volume = {21}, number = {suppl_1}, pages = {i47-i56}, year = {2005}, month = {06}, issn = {1367-4803}, doi = {10.1093/bioinformatics/bti1007}, url = {https://doi.org/10.1093/bioinformatics/bti1007}, eprint = {https://academic.oup.com/bioinformatics/article-pdf/21/suppl_1/i47/524364/bti1007.pdf}, }
贡献者
- 贡献者: @clefourrier




