Omics · study · 2026
SENTRY: Machine Learning Detection of Secondary Senescence
Listed in NCBI GEO
Cellular senescence is a state of irreversible cell cycle arrest that contributes to age-associated decline through the accumulation of senescent cells and their senescence-associated secretory phenotype (SASP).
Description
The SASP, comprising inflammatory signaling molecules and growth factors, can induce secondary senescence in surrounding healthy cells via paracrine signaling. Additionally, secondary senescence can be induced through juxtacrine signaling via activation of NOTCH1.
To further understand the progression of cells into primary and secondary senescence, we used a comprehensive single-cell RNA sequencing data set of clonal human lung fibroblasts (LF1) cell lines from different forms of senescence and quiescence. Here, we present SENTRY (SENescent TRacking sYstem), a new method that uses unsupervised clustering techniques to identify subpopulations of cells common to most major forms of senescence, revealing that the RNA profiles of these subpopulations are driven in part by markers associated with secondary senescence.
Read the rest (2 more)
Leveraging this data, we developed machine learning models using random forests to predict senescent status and subtype with exceptionally high accuracy. SHAP (SHapley Additive exPlanations) analysis identified the most informative genes for classification, many of which are unique to our model compared to existing senescent signatures like SenMayo and SenSig. We then used this classification to analyze single-cell RNA sequencing data in a time course of proliferating and senescent human lung fibroblasts.
We observed that primary and secondary senescent cells exhibit distinct transcriptomic and epigenetic profiles, particularly in pathways related to cell cycle regulation, extracellular matrix remodeling and inflammatory signaling. Additionally, we performed a comparative analysis of gene-length-dependent transcription decline (GLTD) at different stages of senescence induction.
Links
Get the data
- GEO FTP directory ftp.ncbi.nlm.nih.gov/geo/series/GSE306nnn/GSE306635 ↗
download · from NCBI GEO
Where it is published
- GEO accession page ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE306635 ↗
landing page · from NCBI GEO
Documentation and papers
- PRJNA1311543 ncbi.nlm.nih.gov/bioproject/PRJNA1311543 ↗
project · from NCBI GEO
Topics
- Stated by source
- Expression profiling by high throughput sequencing · Homo sapiens
- From keywords
- Life Sciences
- Inferred from text
- RNA sequencing 75% · Sequencing 75% · Single-cell RNA sequencing 75%
Provenance · 1 source records, 10 field assertions
| Source | Key | Last seen | Raw |
|---|---|---|---|
| NCBI GEO | GSE306635 | 15 d ago | JSON v1 |
| Field | Assertion | Extractor | Evidence |
|---|---|---|---|
| access_level | source · NCBI GEO | connector:ncbi_geo@1.0.0 | |
| concepts[field].local:field:life-sciences | mapping · NCBI GEO | connector:ncbi_geo@1.0.0 | |
| concepts[method].geo_series_type:expression-profiling-by-high-throughput-sequencing | source · NCBI GEO | connector:ncbi_geo@1.0.0 | /gdstype |
| concepts[modality].local:modality:rna-seq | enrichment · NCBI GEO | keyword-concept-rules@1.0.0 | title+description (75%) |
| concepts[modality].local:modality:sequencing | enrichment · NCBI GEO | keyword-concept-rules@1.0.0 | title+description (75%) |
| concepts[modality].local:modality:single-cell-rna-seq | enrichment · NCBI GEO | keyword-concept-rules@1.0.0 | title+description (75%) |
| concepts[organism].NCBITaxon:9606 | source · NCBI GEO | connector:ncbi_geo@1.0.0 | /taxon |
| description | source · NCBI GEO | connector:ncbi_geo@1.0.0 | /summary |
| publication_date | source · NCBI GEO | connector:ncbi_geo@1.0.0 | |
| title | source · NCBI GEO | connector:ncbi_geo@1.0.0 | /title |