Back

Are Different Populations Fairly Represented in Single-Cell Omic Atlases?

Yang, C.; Saravanan, K.; Saharan, A.; Huang, K.-l.

2025-10-03 genomics
10.1101/2025.10.01.677375 bioRxiv
Show abstract

Single-cell Omic atlases are transforming biology and medicine, yet their demographic representativeness has not been systematically evaluated. We conducted a secondary analysis of >13,500 samples from the Human Cell Atlas (HCA), Human Tumor Atlas Network (HTAN), and PsychAD Consortium. Benchmarking against global and US general and disease-prevalence data, we found a striking, pervasive European over-representation and underrepresentation of Asian and Latino individuals. Nearly 70% of HCA samples lacked ancestry annotation, HTAN tumors were 69% European, and PsychAD showed more balanced non-European representation but markedly few Asians. Sex distributions in these atlases also skewed compared to published disease prevalence. These disparities highlight that current single-cell resources risk embedding inequities into AI foundational models, biomarker discovery, and therapeutic development.

Published in Cell Genomics (predicted rank #2) · training set

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 12%
15.2%
Similar papers in this journal
Cell Genomics · published here
○ 172 papers in training set
Top 0.3%
7.9%
Similar papers in this journal
3
Scientific Data
○ 209 papers in training set
Top 0.3%
7.9%
Similar papers in this journal
4
The American Journal of Human Genetics
○ 234 papers in training set
Top 0.7%
7.3%
Similar papers in this journal
5
Nature Medicine
○ 125 papers in training set
Top 0.2%
6.8%
Similar papers in this journal
6
Nature
○ 645 papers in training set
Top 3%
5.5%
50% of probability mass above
7
Genome Medicine
○ 183 papers in training set
Top 0.7%
4.9%
8
Human Genetics and Genomics Advances
○ 84 papers in training set
Top 0.5%
4.1%
9
Nature Genetics
○ 286 papers in training set
Top 2%
3.3%
10
Cell
○ 431 papers in training set
Top 3%
3.3%
11
Science
● 477 papers in training set
Top 3%
3.2%
12
Nature Aging
○ 60 papers in training set
Top 0.6%
2.8%
13
Nature Neuroscience
○ 252 papers in training set
Top 3%
2.1%
14
Scientific Reports
○ 3612 papers in training set
Top 51%
1.7%
15
Genetics in Medicine
○ 78 papers in training set
Top 0.6%
1.7%
16
eLife
● 5828 papers in training set
Top 50%
1.7%
17
Alzheimer's & Dementia
177 papers in training set
Top 2%
1.7%
18
Communications Biology
○ 993 papers in training set
Top 21%
1.1%
19
Nature Biotechnology
○ 172 papers in training set
Top 3%
1.1%
20
Genome Biology
○ 637 papers in training set
Top 8%
1.1%
21
Nature Cancer
○ 39 papers in training set
Top 1%
0.9%
22
Nature Human Behaviour
○ 95 papers in training set
Top 2%
0.9%
23
Acta Neuropathologica
○ 58 papers in training set
Top 2%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.