tahoebio/parse-de-rhaister
收藏资源简介:
Parse PBMC细胞因子筛选差异表达汇总统计数据,包括12个供体、18种免疫细胞类型和90种细胞因子的差异表达分析。该数据用于训练Rhaister模型,这些统计量是从Arc Institute的ST-HVG-Parse数据集的原始单细胞中计算得出的,包含每个基因的折叠变化、显著性水平和伪批量差异。细胞上下文是(供体,细胞类型)组合,扰动是细胞因子。数据集包含两个表(pdex和cell_eval)以及定义文件(过滤/拆分标准)。pdex表为长格式,每行对应(供体,细胞类型,细胞因子,基因),约7.74亿行,包含折叠变化、p值等列;cell_eval表为宽格式,每行对应(细胞类型,供体,细胞因子),每列对应一个基因的伪批量差异。
Differential-expression summary statistics for the Parse PBMC cytokine screen including 12 donors × 18 immune cell types × 90 cytokines. This data was used to train Rhaister. These are the statistics computed from the raw single cells of the Arc Institute ST-HVG-Parse dataset, including per-gene fold changes, significance, and pseudobulk deltas. The cell context is a (donor, cell type) combination and the perturbation is a cytokine. The repo contains two tables (pdex and cell_eval) plus the dataset-definition files (filtering / split criteria).




