Gene markers of senescence vary in reliability in human cells
A computational analysis compared nine gene sets across 25 datasets, six human cell types and 12 senescence-inducing conditions, finding wide differences in performance.
Our read
AI & data- Studied:
- Computational study of human cells, against dividing and resting cells
- What changed:
- Senescence gene markers varied in reliability: SenMayo performed reliably across contexts, while MSigDB sets performed poorly
- The catch:
- Differences between tissues may reflect cell-type mixtures rather than senescence levels
- In the record:
- One of 5 papers on senescence-related gene signatures on record: 3 with results in people, 1 in animals.
Editor of record: Boris Djordjevic · AgingResearch.News
A computational study of six types of human cells compared gene activity markers of senescence, a state in which cells stop dividing. The researchers developed markeR, a freely available software package for evaluating groups of genes as markers of cell states. They tested nine published gene sets across 25 datasets and 12 conditions that induced senescence.
Performance varied widely. SenMayo, one of the gene sets, performed reliably across contexts, while some commonly used alternatives performed poorly. The analysis included comparisons with dividing cells and with resting cells that had also stopped dividing. Separately, the researchers applied the software to data from 49 human tissue types. They found tissue- and age-related differences in gene activity signals associated with senescence.
Why it matters
Senescence has been linked to multiple diseases. Reliable molecular markers could help researchers investigate how this cell state relates to tissue aging and clinical features.
Caveats
Gene activity signals are indirect markers, and the tissue analysis lacked a definitive reference measure of senescence. Differences between tissues may reflect their mix of cell types rather than different levels of senescence.
- Resources: datasets, atlases and tools
- Cellular senescence
- SenMayo
- MSigDB gene sets
- Senescence-related gene signatures
- Humans
The paper
University of Lisbon
NAR Genomics and Bioinformatics, Mar 27, 2026, Open access
Comments
Be the first to comment.