Planned analysis: integrate Allen Immunology aging cohort scRNA-seq with Mogilenko 2021 and Terekhova 2023 CITE-seq datasets. Pseudobulk per donor per CD8 subset (limma-voom, covariates: age-decade, sex, CMV status, cohort). Score each donor-level pseudobulk for (1) Tex module (TOX, PDCD1, LAG3, HAVCR2, TIGIT, TCF7-low) and (2) senescence module (B3GAT1/CD57, KLRG1, GZMB, SELL-low, IL7R-low). Test decade × module interaction. Success criterion: FDR-adjusted decade effect on Tex-vs-senescence module ratio surviving CMV-status adjustment in ≥2 independent cohorts. Datasets required: Allen cohort (internal), GSE accessions for Mogilenko 2021 and Terekhova 2023 (TODO_VERIFY). Wells et al. 2025 (Nat Immunol, DOI 10.1038/s41590-025-02241-4) flagged as tissue-compartment reference for cross-tissue replication arm.
Details
- disease
- immune aging / immunosenescence
- target_ref
- CD8A, PDCD1, TOX, B3GAT1, KLRG1, GZMK, GZMB
- primary_endpoint
- Gate PASS: Wells 2025 full text yields ≥3 of 5 separability fields (donor-n with age range, effect size, p/FDR, CMV status, TCR clonotype) for GZMK+ CD8 in ≥2 non-blood tissues
- identification_strategy
- observational
Raw fields (3)
- assay_spec
Literature triangulation: extract quantitative CD8 TEM exhaustion-vs-senescence separability metrics from Wells et al. 2025 (DOI 10.1038/s41590-025-02241-4) full text via PubMed/PMC fetch, cross-referenced against Mogilenko 2021 GZMK+ clonal expansion data and Terekhova 2023 NKG2C+CD57+ loss metrics, then score a pass/fail gate on five pre-specified separability fields; Separability gate scorecard: five fields (donor n with age range, effect size R2 or logFC, p-value or FDR, CMV status reported Y/N, TCR clonotype data Y/N) scored present/absent per paper; gate PASS if Wells 2025 provides ≥3 of 5 fields across ≥2 tissues; {'kind': 'claim', 'text_template': 'Wells et al. 2025 [PASS/FAIL/BLOCKED]: GZMK+ CD8 tissue-directed aging signature reports [n] of 5 required separability fields; CMV-adjusted analysis [present/absent]; TCR clonotype overlay [present/absent]; tissues covered: [list]'}; Named blocker statement if gate fails: exact missing field(s) and which dataset or analysis step would supply them, expressed as a concrete next-tick fetch or analysis target- hypothesis
In aged donors (>60y), circulating and tissue-resident CD8 T cells harbour transcriptionally and phenotypically distinct exhausted (TOX+PD-1+LAG3+) versus senescent (CD57+KLRG1+GZMB+IL7R-low) subsets that contribute independently to the GZMK+ inflammaging signature; their relative abundance is confounded by CMV serostatus and sex, and requires per-donor pseudobulk modelling with these covariates to partition correctly.
- kill_criteria
Abort or revise if: Wells 2025 PMC full text may not include per-donor CMV serostatus — likely BLOCKED on that field given tissue atlas scope; Mogilenko 2021 CMV reporting is known to be present but may not include TCR clonotype overlap with tissue data; Notebook cells 100-106 may already contain partial extractions; must read before re-extracting to avoid duplicate claims