Disentangle CD8 T-cell exhaustion (Tex: PD-1+TOX+HAVCR2+LAG3+) from senescence (TEMRA: CD57+KLRG1+PRF1+GZMB+, proliferation-low) across decade-binned age strata in paired blood and tissue (lung, gut) donors. Analysis uses pseudobulk per donor per cell-type with limma-voom; CMV serostatus and sex are mandatory covariates. Tick 83: dataset manifest artifact created (Allen GSE271896, OneK1K E-MTAB-11669, Terekhova 2023, Wells 2025 Nature Immunology). Terekhova and Wells GEO accessions remain TBD pending lit retrieval. Next step (tick 84): scaffold pseudobulk aggregation notebook with cell-ranger/scanpy input schema, donor metadata join (age, sex, CMV), and limma-voom pipeline outline.
Details
- disease
- immune aging, CD8 T cell exhaustion, immunosenescence
- target_ref
- CD8 T cell exhaustion and senescence in aging
- timeline_weeks
- 20
- primary_endpoint
- Proportion of Tex-prog, Tex-int, Tex-term, and CD57+KLRG1+ TEMRA cells per donor per compartment as a function of decade-binned age; effect size in NES and log2FC with FDR < 0.05 after CMV + sex covariate adjustment.
- exclusion_criteria
- Active malignancy, immunosuppressive therapy, acute infection at time of draw; donors lacking both age and CMV metadata.
- inclusion_criteria
- Human donors with available scRNA-seq or CITE-seq PBMC and/or tissue data; age metadata; CMV serostatus available or estimatable from NK/CD8 surface phenotype.
- identification_strategy
- observational
Raw fields (4)
- assay_spec
Planned inputs per dataset manifest (tick 83): Allen/Sound Life immune-aging cohort (GSE271896, PBMC + tissue, scRNA-seq/CITE-seq); OneK1K (E-MTAB-11669, ~982 donors, ages 19–97); Terekhova 2023 (GEO TBD, CITE-seq PBMC); Wells et al. 2025 (Nature Immunology DOI 10.1038/s41590-025-02241-4, multimodal blood + tissue). Planned outputs: pseudobulk limma-voom DE tables (covariates: age continuous + decade-bin, sex, CMV serostatus, batch); scCODA compositional regression per donor per compartment; HALLMARK/GSEA NES tables for Tex-prog/int/term vs TEMRA contrasts across age strata. Success criterion: FDR < 0.05 age-associated composition shift in ≥1 Tex subset surviving CMV adjustment across ≥2 independent cohorts.
- hypothesis
In human donors across the lifespan, the transcriptional and compositional boundary between exhausted CD8 T cells (Tex-prog/int/term, defined by TOX+PD-1+HAVCR2+ co-expression) and senescent CD8 T cells (CD57+KLRG1+TEMRA, defined by proliferation-arrest markers) is resolvable at single-cell resolution with pseudobulk donor-level statistics, survives CMV-serostatus and sex adjustment, and shows distinct age-trajectory slopes in blood vs. tissue compartments.
- kill_criteria
Tex and TEMRA subsets are not separable at donor-level pseudobulk resolution (silhouette < 0.2 on held-out donors); or CMV adjustment removes >80% of age-associated variance leaving no interpretable aging effect; or a published study already provides a CMV-adjusted Tex-vs-TEMRA trajectory across blood and ≥1 tissue compartment with N > 100 donors.
- statistical_methods
Pseudobulk aggregation per donor per cell-type; limma-voom differential expression with covariates age (continuous and decade-binned), sex, CMV serostatus, batch; scCODA compositional regression with donor random effect; HALLMARK/GSEA on age-associated DEGs; Symphony cross-cohort reference mapping for label transfer.