tahoebio/tahoe-de-rhaister
收藏资源简介:
Tahoe-100M药物筛选差异表达汇总统计数据,包含50个癌细胞系×384种药物×3个剂量的差异表达摘要,是Rhaister模型的训练数据集之一。这些统计数据基于约1亿个原始单细胞(来自tahoebio/Tahoe-100M数据集)计算得出,包括每基因的折叠变化、显著性以及每个(细胞系,处理)的伪批量差异。数据集包含两个表格(pdex和cell_eval)以及用于零样本评估的对照状态参考,还有数据集定义文件(过滤/拆分标准)。
Differential-expression summary statistics for the Tahoe-100M drug screen including 50 cancer cell lines × 384 drugs × 3 doses, one of the training datasets for Rhaister. These are the statistics computed from the ~100M raw single cells of tahoebio/Tahoe-100M, including per-gene fold changes, significance, and pseudobulk deltas per (cell line, treatment). The repo includes two tables (pdex and cell_eval) plus the control-state references used for zero-shot evaluation, and the dataset-definition files (filtering / split criteria).




