Back

A single-cell atlas and aging clock define biological age and risk-associated stem cell states in human hematopoiesis

Chen, H.; Dong, P.; Xu, J.; Wang, G.

2026-02-13 bioinformatics
10.64898/2026.02.12.703707 bioRxiv
Show abstract

Aging of hematopoietic stem and progenitor cells (HSPCs) impairs regenerative capacity and predisposes to hematological diseases. Here, we constructed a comprehensive single-cell transcriptomic atlas comprising 186,123 CD34+ HSPCs spanning early prenatal development (6 post-conception weeks) to late adulthood (74 years). We identified two conserved core molecular programs (MPs) of inflammaging and RNA splicing / protein homeostasis. Leveraging these programs, we developed a machine learning-based stem cell aging clock from 84 donors. Applying this clock to acute myeloid leukemia (AML), we define Transcriptional Age Deviation (TAD), a novel metric of biological age divergence. We found that a biologically "younger" state (low TAD), reflecting oncofetal reprogramming, is a powerful independent predictor of poor survival in two large AML cohorts. Low TAD was associated with high-risk genetics and therapy resistance, and critically, it re-stratified patient outcomes within each ELN 2022 risk category. Our work establishes a quantitative link between stem cell aging biology and AML prognosis, offering a robust tool to refine clinical risk assessment.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 4%
26.4%
Similar papers in this journal
2
Nature
○ 645 papers in training set
Top 3%
5.5%
Similar papers in this journal
3
Cell Reports
○ 1498 papers in training set
Top 6%
5.4%
Similar papers in this journal
4
Science Advances
● 1243 papers in training set
Top 5%
5.1%
Similar papers in this journal
5
Advanced Science
○ 286 papers in training set
Top 2%
4.0%
Similar papers in this journal
6
Leukemia
○ 42 papers in training set
Top 0.3%
3.4%
7
Cell Research
○ 51 papers in training set
Top 0.2%
3.1%
50% of probability mass above
8
eLife
● 5828 papers in training set
Top 42%
2.4%
9
Cell Genomics
○ 172 papers in training set
Top 2%
2.4%
10
Nature Aging
○ 60 papers in training set
Top 0.8%
2.1%
11
Stem Cell Reports
○ 130 papers in training set
Top 1%
2.1%
12
Science
● 477 papers in training set
Top 4%
2.1%
13
Nature Genetics
○ 286 papers in training set
Top 3%
2.1%
14
Communications Biology
○ 993 papers in training set
Top 14%
1.7%
15
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 29%
1.7%
16
Cell Stem Cell
○ 62 papers in training set
Top 1%
1.5%
17
Cell Reports Medicine
○ 153 papers in training set
Top 3%
1.5%
18
Genome Biology
○ 637 papers in training set
Top 6%
1.5%
19
Nature Cell Biology
○ 118 papers in training set
Top 2%
1.5%
20
iScience
○ 1154 papers in training set
Top 22%
1.4%
21
Genome Medicine
○ 183 papers in training set
Top 3%
1.3%
22
Nucleic Acids Research
◐ 1281 papers in training set
Top 11%
1.1%
23
Cell Systems
○ 201 papers in training set
Top 4%
1.1%
24
Nature Metabolism
○ 69 papers in training set
Top 2%
1.1%
25
Briefings in Bioinformatics
◐ 354 papers in training set
Top 6%
1.0%
26
Cell
○ 431 papers in training set
Top 10%
0.8%
27
Nature Biomedical Engineering
○ 47 papers in training set
Top 2%
0.6%
28
Autophagy
○ 39 papers in training set
Top 0.7%
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.