Evaluating the influence of structural properties on proximity metric performance in single cell RNA-seq data - Datasets

Includes raw and processed copies of the scRNA-seq datasets used for the paper: 'How does data structure impact cell-cell similarity? Evaluating the influence of structural properties on proximity metric performance in single cell RNA-seq data.'

Real scRNA-seq.zip contains the Abundant (subset1) and Rare (subset 2) subsets generated to represent discretely structured datasets (sourced from Wegmann et al. 2019) and the continuously structured data (sourced from Popescu et al. 2019).

Simulated scRNA-seq.zip contains the Abundant, Moderately-Rare and Ultra-Rare subsets for discretely and continuously structured datasets. All data was simulated using the PROSSTT package in Python 3.8, as well as the dataset containing the labels to re-produce Figure 3 of the manuscript.

Results.zip contains the results for all datasets from the full analysis, in a pickled python dictionary. Code to read in and visualise results is available on the projects github

The scripts for the dataset generation, processing and visualisation of results are available at our github for the scProcimitE package, and documentation is available here.

本数据集包含本论文《数据结构如何影响细胞间相似性？评估单细胞RNA测序数据中结构特性对邻近指标性能的影响》所使用的原始与预处理后的单细胞RNA测序（scRNA-seq）数据集副本。

Real scRNA-seq.zip 包含用于表征离散结构数据集的丰度（子集1）与稀有（子集2）子集（数据源自Wegmann等，2019年研究），以及连续结构数据集（数据源自Popescu等，2019年研究）。

Simulated scRNA-seq.zip 包含离散与连续结构数据集对应的丰度、中等稀有与超稀有子集。所有数据均使用Python 3.8环境下的PROSSTT工具包生成，同时附带可复现论文图3的带标签数据集。

Results.zip 包含全部数据集完整分析后的结果，以Python序列化字典格式存储。用于读取与可视化结果的代码可在本项目的GitHub仓库中获取。

用于数据集生成、预处理与结果可视化的脚本，以及scProcimitE软件包的相关文档，均可在本团队的GitHub仓库中获取，文档链接亦可在此处查阅。