Dataset Viewer
Auto-converted to Parquet Duplicate
gene
stringlengths
1
10
compound
stringlengths
13
21
logFC
float32
-10.82
24.3
⌀
pvalue
float32
0
1
⌀
A2M
MCE25_293T_24H_X2:A03
-0.282386
0.901582
A2M
MCE25_293T_24H_X2:A04
0.93047
0.646084
A2M
MCE25_293T_24H_X2:A05
-0.899633
0.692461
A2M
MCE25_293T_24H_X2:A06
3.162598
0.023761
A2M
MCE25_293T_24H_X2:A07
1.097357
0.562184
A2M
MCE25_293T_24H_X2:A08
-0.804043
0.794099
A2M
MCE25_293T_24H_X2:A09
-0.494561
0.826445
A2M
MCE25_293T_24H_X2:A10
0.048926
0.98026
A2M
MCE25_293T_24H_X2:A11
-0.593118
0.7949
A2M
MCE25_293T_24H_X2:A12
0.39358
0.84095
A2M
MCE25_293T_24H_X2:A13
-0.87204
0.741379
A2M
MCE25_293T_24H_X2:A14
1.580589
0.358829
A2M
MCE25_293T_24H_X2:A16
-0.26367
0.90722
A2M
MCE25_293T_24H_X2:A17
-0.418303
0.874992
A2M
MCE25_293T_24H_X2:A18
0.87923
0.659108
A2M
MCE25_293T_24H_X2:A19
-0.992357
0.711094
A2M
MCE25_293T_24H_X2:A21
-0.363405
0.89084
A2M
MCE25_293T_24H_X2:A22
0.783811
0.693191
A2M
MCE25_293T_24H_X2:B03
1.587345
0.39632
A2M
MCE25_293T_24H_X2:B04
-0.661892
0.803638
A2M
MCE25_293T_24H_X2:B05
-0.071767
0.971075
A2M
MCE25_293T_24H_X2:B06
-0.234704
0.933927
A2M
MCE25_293T_24H_X2:B07
-0.910758
0.731973
A2M
MCE25_293T_24H_X2:B08
-1.003933
0.661638
A2M
MCE25_293T_24H_X2:B09
0.354634
0.879561
A2M
MCE25_293T_24H_X2:B10
0.211036
0.916252
A2M
MCE25_293T_24H_X2:B11
1.244669
0.648354
A2M
MCE25_293T_24H_X2:B12
-0.79227
0.767391
A2M
MCE25_293T_24H_X2:B13
-1.162382
0.603189
A2M
MCE25_293T_24H_X2:B14
-0.556123
0.803448
A2M
MCE25_293T_24H_X2:B15
-0.723861
0.751422
A2M
MCE25_293T_24H_X2:B17
-1.095536
0.662342
A2M
MCE25_293T_24H_X2:B18
0.651302
0.743096
A2M
MCE25_293T_24H_X2:B19
-1.039986
0.646594
A2M
MCE25_293T_24H_X2:B20
-1.141359
0.610733
A2M
MCE25_293T_24H_X2:B21
1.210334
0.455375
A2M
MCE25_293T_24H_X2:B22
-1.087853
0.694078
A2M
MCE25_293T_24H_X2:C03
-0.149363
0.948771
A2M
MCE25_293T_24H_X2:C05
-0.612424
0.789349
A2M
MCE25_293T_24H_X2:C06
-0.697515
0.757959
A2M
MCE25_293T_24H_X2:C07
0.754262
0.701571
A2M
MCE25_293T_24H_X2:C08
-0.59778
0.788335
A2M
MCE25_293T_24H_X2:C09
-0.757077
0.735406
A2M
MCE25_293T_24H_X2:C10
-0.308719
0.892982
A2M
MCE25_293T_24H_X2:C11
1.109412
0.570837
A2M
MCE25_293T_24H_X2:C12
0.66375
0.736564
A2M
MCE25_293T_24H_X2:C13
-0.97275
0.721423
A2M
MCE25_293T_24H_X2:C14
-0.960281
0.673227
A2M
MCE25_293T_24H_X2:C16
-0.166101
0.943384
A2M
MCE25_293T_24H_X2:C17
-0.946682
0.680653
A2M
MCE25_293T_24H_X2:C18
-0.617867
0.780382
A2M
MCE25_293T_24H_X2:C19
-0.935236
0.67996
A2M
MCE25_293T_24H_X2:C20
-0.438074
0.847935
A2M
MCE25_293T_24H_X2:C21
-0.365089
0.871273
A2M
MCE25_293T_24H_X2:D03
0.327054
0.909078
A2M
MCE25_293T_24H_X2:D04
1.149996
0.495671
A2M
MCE25_293T_24H_X2:D06
-0.8412
0.720065
A2M
MCE25_293T_24H_X2:D07
1.092609
0.632946
A2M
MCE25_293T_24H_X2:D08
-0.852674
0.752713
A2M
MCE25_293T_24H_X2:D09
-0.795655
0.72481
A2M
MCE25_293T_24H_X2:D10
1.168807
0.603442
A2M
MCE25_293T_24H_X2:D11
-0.844687
0.753815
A2M
MCE25_293T_24H_X2:D12
-0.725955
0.749034
A2M
MCE25_293T_24H_X2:D13
0.595789
0.763332
A2M
MCE25_293T_24H_X2:D14
-0.224401
0.935734
A2M
MCE25_293T_24H_X2:D15
-0.780179
0.729298
A2M
MCE25_293T_24H_X2:D16
-0.207757
0.926515
A2M
MCE25_293T_24H_X2:D17
1.189113
0.527114
A2M
MCE25_293T_24H_X2:D18
0.32888
0.867395
A2M
MCE25_293T_24H_X2:D19
-1.137258
0.656492
A2M
MCE25_293T_24H_X2:D21
-0.949596
0.673735
A2M
MCE25_293T_24H_X2:D22
1.023172
0.545936
A2M
MCE25_293T_24H_X2:E04
-1.183559
0.595675
A2M
MCE25_293T_24H_X2:E05
-0.963624
0.667505
A2M
MCE25_293T_24H_X2:E06
-0.980131
0.710689
A2M
MCE25_293T_24H_X2:E08
1.686026
0.288789
A2M
MCE25_293T_24H_X2:E09
-0.507592
0.820769
A2M
MCE25_293T_24H_X2:E10
-0.487096
0.854115
A2M
MCE25_293T_24H_X2:E11
-0.547776
0.810575
A2M
MCE25_293T_24H_X2:E12
0.853996
0.661251
A2M
MCE25_293T_24H_X2:E14
0.066058
0.973485
A2M
MCE25_293T_24H_X2:E15
-0.319956
0.886958
A2M
MCE25_293T_24H_X2:E16
-0.645418
0.775646
A2M
MCE25_293T_24H_X2:E17
-0.366967
0.871921
A2M
MCE25_293T_24H_X2:E19
2.142798
0.404977
A2M
MCE25_293T_24H_X2:E20
-0.417882
0.874313
A2M
MCE25_293T_24H_X2:E21
0.955926
0.613765
A2M
MCE25_293T_24H_X2:E22
-0.563443
0.802528
A2M
MCE25_293T_24H_X2:F03
-0.414778
0.854411
A2M
MCE25_293T_24H_X2:F04
-0.956037
0.669857
A2M
MCE25_293T_24H_X2:F05
1.016214
0.660569
A2M
MCE25_293T_24H_X2:F06
-0.615845
0.786188
A2M
MCE25_293T_24H_X2:F08
-0.186515
0.942027
A2M
MCE25_293T_24H_X2:F09
-0.891108
0.695339
A2M
MCE25_293T_24H_X2:F10
0.318983
0.8735
A2M
MCE25_293T_24H_X2:F11
-0.505303
0.829337
A2M
MCE25_293T_24H_X2:F12
0.779179
0.692604
A2M
MCE25_293T_24H_X2:F13
-0.198716
0.931656
A2M
MCE25_293T_24H_X2:F14
1.388317
0.4788
A2M
MCE25_293T_24H_X2:F15
1.030575
0.544879
End of preview. Expand in Data Studio

CP-TCM:化合物扰动下的基因 logFC / p 值矩阵

数据集概述

本数据集记录 27,548 个化合物组合在 3,407 个基因上的扰动响应,提供两套数值:

子集(Subsets)

config_name 内容 规模
expression_data 长表:gene / compound / logFC / pvalue 93,856,036 行
compound_metadata 化合物注释(一行一个 compound) 27,548 行
gene_metadata 基因注释(一行一个 gene) 3,407 行

在 Dataset Viewer 左上角的 Subset 下拉框里切换即可分别浏览与查看各列的统计分布。

字段说明

expression_data

列 类型 含义
gene string 基因名
compound string 化合物/组合名
logFC float32 相对对照的 log2 fold change
pvalue float32 原始 p 值(未做多重校正)

compound_metadata

compound(主键)+ 原始 meta 表中的各注释列(靶点、剂量、批次等)。

gene_metadata

列 类型 含义
gene string 基因名(主键)
gene_index int32 在原矩阵中的行序号

目录结构

.
├── README.md
├── data/
│   ├── compound_00000.parquet      # gene, compound, logFC, pvalue
│   ├── compound_00001.parquet
│   └── ...
└── meta/
    ├── compounds.parquet           # 一行一个 compound,含化合物注释列
    └── genes.parquet               # 一行一个 gene

快速使用

from datasets import load_dataset

# 只取一部分(推荐,避免一次拉 1 GB)
ds = load_dataset("your-name/cp-tcm", split="train", streaming=True)
for row in ds.take(10):
    print(row)
import pyarrow.dataset as pads

d = pads.dataset("cp_tcm_hf/data", format="parquet")
tbl = d.to_table(filter=(pads.field("compound") == "某个化合物名"))
print(tbl.to_pandas())
# 取回宽矩阵(基因 × 化合物)
import pandas as pd, glob
df = pd.concat([pd.read_parquet(f) for f in sorted(glob.glob("data/*.parquet"))])
wide = df.pivot(index="gene", columns="compound", values="logFC")

数据处理说明

  • 原始文件为两个 CSV(cp_tcm_fc.csv、cp_tcm_pval.csv)与一个 meta 表;
  • 转换为长表 Parquet(zstd 压缩,float32),按 500 个化合物一片;
  • meta/compounds.parquet 的行顺序与 data/ 分片的化合物顺序一致,便于按序重建宽矩阵;
  • 行顺序:分片内先按 gene、再按 compound 排列(gene 变化最慢)。

已知限制

  • pvalue 为原始 p 值,未做 BH/FDR 校正,跨化合物比较请自行校正;
  • logFC 的对照定义、归一化方式与原实验一致(见源文献);
  • 化合物名中的特殊字符(,、(、+)已原样保留,读取时请勿自行替换。

引用

如使用本数据集,请引用:<原始论文/数据来源>。

License

<许可证说明>

Downloads last month
-