Yunnie0216/ZINC

--- license: unknown dataset_info: features: - name: node_feat sequence: sequence: int64 - name: edge_index sequence: sequence: int64 - name: edge_attr sequence: sequence: int64 - name: 'y' sequence: float64 - name: num_nodes dtype: int64 splits: - name: train num_bytes: 376796456 num_examples: 220011 - name: test num_bytes: 8538528 num_examples: 5000 - name: validation num_bytes: 41819628 num_examples: 24445 download_size: 20636253 dataset_size: 427154612 task_categories: - graph-ml --- # Dataset Card for ZINC ## Table of Contents - [Table of Contents](#table-of-contents) - [Dataset Description](#dataset-description) - [Dataset Summary](#dataset-summary) - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards) - [External Use](#external-use) - [PyGeometric](#pygeometric) - [Dataset Structure](#dataset-structure) - [Data Properties](#data-properties) - [Data Fields](#data-fields) - [Data Splits](#data-splits) - [Additional Information](#additional-information) - [Licensing Information](#licensing-information) - [Citation Information](#citation-information) - [Contributions](#contributions) ## Dataset Description - **[Homepage](https://zinc15.docking.org/)** - **[Repository](https://www.dropbox.com/s/feo9qle74kg48gy/molecules.zip?dl=1):**: - **Paper:**: ZINC 15 – Ligand Discovery for Everyone (see citation) - **Leaderboard:**: [Papers with code leaderboard](https://paperswithcode.com/sota/) ### Dataset Summary The `ZINC` dataset is a "curated collection of commercially available chemical compounds prepared especially for virtual screening" (Wikipedia). ### Supported Tasks and Leaderboards `ZINC` should be used for molecular property prediction (aiming to predict the constrained solubility of the molecules), a graph regression task. The score used is the MAE. The associated leaderboard is here: [Papers with code leaderboard](https://paperswithcode.com/sota/graph-regression-on-zinc). ## External Use ### PyGeometric To load in PyGeometric, do the following: ```python from datasets import load_dataset from torch_geometric.data import Data from torch_geometric.loader import DataLoader dataset_hf = load_dataset("graphs-datasets/<mydataset>") # For the train set (replace by valid or test as needed) dataset_pg_list = [Data(graph) for graph in dataset_hf["train"]] dataset_pg = DataLoader(dataset_pg_list) ``` ## Dataset Structure ### Data Properties | property | value | |---|---| | scale | big | | #graphs | 220011 | | average #nodes | 23.15 | | average #edges | 49.81 | ### Data Fields Each row of a given file is a graph, with: - `node_feat` (list: #nodes x #node-features): nodes - `edge_index` (list: 2 x #edges): pairs of nodes constituting edges - `edge_attr` (list: #edges x #edge-features): for the aforementioned edges, contains their features - `y` (list: 1 x #labels): contains the number of labels available to predict (here 1, equal to zero or one) - `num_nodes` (int): number of nodes of the graph ### Data Splits This data comes from the PyGeometric version of the dataset, and follows the provided data splits. This information can be found back using ```python from torch_geometric.datasets import ZINC dataset = ZINC(root = '', split='train') # valid, test ``` ## Additional Information ### Licensing Information The dataset has been released under unknown license. Please open an issue if you know what is the license of this dataset. ### Citation Information ```bibtex @article{doi:10.1021/acs.jcim.5b00559, author = {Sterling, Teague and Irwin, John J.}, title = {ZINC 15 – Ligand Discovery for Everyone}, journal = {Journal of Chemical Information and Modeling}, volume = {55}, number = {11}, pages = {2324-2337}, year = {2015}, doi = {10.1021/acs.jcim.5b00559}, note ={PMID: 26479676}, URL = { https://doi.org/10.1021/acs.jcim.5b00559 }, eprint = { https://doi.org/10.1021/acs.jcim.5b00559 } } ``` ### Contributions Thanks to [@clefourrier](https://github.com/clefourrier) for adding this dataset.license: 未知 dataset_info: features: - name: `node_feat`（节点特征） sequence: sequence: int64 - name: `edge_index`（边索引） sequence: sequence: int64 - name: `edge_attr`（边属性） sequence: sequence: int64 - name: 'y'（标签） sequence: float64 - name: num_nodes（节点数量） dtype: int64 splits: - name: 训练集（train） num_bytes: 376796456 num_examples: 220011 - name: 测试集（test） num_bytes: 8538528 num_examples: 5000 - name: 验证集（validation） num_bytes: 41819628 num_examples: 24445 download_size: 20636253 dataset_size: 427154612 task_categories: - 图机器学习（graph-ml） --- # ZINC数据集卡片 ## 目录 - [目录](#目录) - [数据集描述](#数据集描述) - [数据集概述](#数据集概述) - [支持任务与排行榜](#支持任务与排行榜) - [外部使用](#外部使用) - [PyTorch几何（PyGeometric）](#pytorch几何pygeometric) - [数据集结构](#数据集结构) - [数据属性](#数据属性) - [数据字段](#数据字段) - [数据划分](#数据划分) - [附加信息](#附加信息) - [许可证信息](#许可证信息) - [引用信息](#引用信息) - [贡献者](#贡献者) ## 数据集描述 - **[主页](https://zinc15.docking.org/)** - **[仓库](https://www.dropbox.com/s/feo9qle74kg48gy/molecules.zip?dl=1):** - **论文:** ZINC 15 – 面向大众的配体发现（详见引用信息） - **排行榜:** [Papers with code 排行榜](https://paperswithcode.com/sota/) ### 数据集概述 ZINC数据集是"专为虚拟筛选精心整理的商用化学化合物合集"（引自维基百科）。 ### 支持任务与排行榜 ZINC数据集适用于分子属性预测任务（目标为预测分子的受限溶解度），属于图回归任务。评估指标采用平均绝对误差（MAE，Mean Absolute Error）。 相关排行榜链接：[Papers with code 排行榜](https://paperswithcode.com/sota/graph-regression-on-zinc) ## 外部使用 ### PyTorch几何（PyGeometric） 若需在PyGeometric中加载该数据集，可执行如下代码： python from datasets import load_dataset from torch_geometric.data import Data from torch_geometric.loader import DataLoader dataset_hf = load_dataset("graphs-datasets/<mydataset>") # 针对训练集（可根据需要替换为验证集或测试集） dataset_pg_list = [Data(graph) for graph in dataset_hf["train"]] dataset_pg = DataLoader(dataset_pg_list) ## 数据集结构 ### 数据属性 | 属性 | 取值 | |---|---| | 规模 | 大型 | | 图总数 | 220011 | | 平均节点数 | 23.15 | | 平均边数 | 49.81 | ### 数据字段 每个文件的每一行对应一张图，包含以下字段： - `node_feat`（节点特征）：维度为#节点 × #节点特征的矩阵，存储节点特征 - `edge_index`（边索引）：维度为2 × #边数的矩阵，存储构成边的节点对 - `edge_attr`（边属性）：维度为#边数 × #边特征的矩阵，存储对应边的特征信息 - `y`（标签）：维度为1 × #标签数的向量，存储待预测的标签（此处仅1个标签，取值为0或1） - `num_nodes`（节点数量）：整数类型，存储该图的总节点数 ### 数据划分 本数据集源自PyGeometric版本的ZINC数据集，采用其官方数据划分规则。可通过以下代码获取划分信息： python from torch_geometric.datasets import ZINC dataset = ZINC(root = '', split='train') # 可替换为valid或test ## 附加信息 ### 许可证信息 本数据集采用未知许可证发布。若您知晓该数据集的具体许可证，请提交Issue告知。 ### 引用信息 bibtex @article{doi:10.1021/acs.jcim.5b00559, author = {Sterling, Teague and Irwin, John J.}, title = {ZINC 15 – 面向大众的配体发现}, journal = {Journal of Chemical Information and Modeling}, volume = {55}, number = {11}, pages = {2324-2337}, year = {2015}, doi = {10.1021/acs.jcim.5b00559}, note ={PMID: 26479676}, URL = { https://doi.org/10.1021/acs.jcim.5b00559 }, eprint = { https://doi.org/10.1021/acs.jcim.5b00559 } } ### 贡献者 感谢 [@clefourrier](https://github.com/clefourrier) 为本数据集的收录提供支持。