Paper Evidence Map: T2D PBMC Ancestry

This synthesis page is the working bridge between source notes and manuscript structure. Do not treat any row as settled until it links to ingested references.

Evidence Matrix

Transclude of paper-evidence-map-t2d-pbmc-ancestry.univer

Topic	Evidence Summary	Key Sources	Confidence	Manuscript Use
PBMC cell composition in T2D	Gu et al. report higher NK cells and CD16 monocytes in T2D and higher B cells and CD14 monocytes in non-diabetes.	Gu et al. 2024	single-study	Background/results framing
NK cell composition in T2D PBMCs varies across studies	Tkachenko et al. 2024 report NK cell depletion in T2D PBMCs, while Gu et al. 2024 observed different NK trends; this illustrates cross-study heterogeneity at the PBMC single-cell level.	Tkachenko et al. 2024, Gu et al. 2024	single-study each; Tkachenko is a preprint	Limitations/replication framing — cross-study variability in PBMC cell composition
Bulk blood T2D transcriptome signal-to-noise	Tkachenko et al. 2024 bulk RNA-seq PCA shows no T2D group separation; only 146/74K genes are DE.	Tkachenko et al. 2024	single-study preprint	Methods/limitations — supports caution about bulk-blood signal-to-noise and motivates testing whether PBMC assays provide clearer cell-type-resolved signals
PBMC transcriptomic changes in T2D	Gu et al. report inflammatory CD14 monocyte states, MHC class II signal in intermediate monocytes, cytotoxic T-cell expansion, and B-cell differentiation/isotype changes.	Gu et al. 2024	single-study	Mechanistic framing
PBMC inflammatory pathway signaling in T2D	Zhao and Fang report T-cell and monocyte DEG/pathway findings in a small Chinese PBMC scRNA-seq cohort, including TNF/NF-kB, interferon-gamma response, T-cell receptor signaling, chemokine signaling, and TNFRSF1A-centered T-cell network interactions.	Zhao and Fang 2025, t2d-pbmc-tnf-ifng-signaling	exploratory single-study, n=3/group	Background/mechanistic framing — supports inflammatory pathway plausibility but not ancestry-specific claims.
Ancestry-stratified PBMC findings	Pending literature intake	Pending	unreviewed	Central contribution
Confounding by BMI, medication, glycemia, or site	Gu et al. note a modest BMI correlation with CD14 monocyte inflammation score and missing medication information for T2D patients.	Gu et al. 2024	single-study	Limitations and interpretation
Replication across cohorts	Pending literature intake	Pending	unreviewed	Strength of evidence
Replication across blood transcriptomic cohorts	Tkachenko et al. found low cross-study concordance in T2D blood RNA-seq DEGs; four of eight datasets had zero DEGs, and only five genes were directionally shared across the three DEG-rich studies.	Tkachenko et al. 2025	multi-study meta-analysis	Limitations/replication framing
Cross-study T2D blood pathways	Meta-analysis identified 2065 DEGs and 713 integration-driven discoveries enriched for neutrophil effector compartments, ERAD, mTOR, oxidative stress, and RNA-splicing themes.	Tkachenko et al. 2025	multi-study meta-analysis	Mechanistic comparison and candidate pathway framing
Whole-blood versus PBMC interpretation	Tkachenko et al. includes six whole-blood and two PBMC datasets, so findings are blood-transcriptomic context rather than strictly PBMC-specific evidence.	Tkachenko et al. 2025	interpreted limitation	Methods/limitations
PBMC immunometabolic heterogeneity in T2D	Li et al. 2025 stratified T2D samples into subtypes A-C from T-cell metabolic pathway GSVA scores and linked subtypes to immune-cell proportions, inferred T-cell-monocyte communication, inferred TF activity, and drug enrichment.	Li et al. 2025	single-study reanalysis	Mechanistic framing and hypothesis generation
Dataset reuse and independence	Li et al. 2025 reuses the GSE268210 T2D samples previously analyzed by Gu et al. 2024; it adds GSE244515 controls but is not an independent T2D replication source.	gse268210, gse244515	interpretation constraint	Limitations and evidence weighting
T2D islet transcriptomic biomarkers	Huang et al. report 4 candidate diagnostic genes (SLC2A2, SERPINF1, RASGRP1, CHL1) from islet RNA-seq, with SERPINF1-NR2F2 regulatory-axis analysis in pancreatic fibroblasts.	Huang et al. 2022	single-study	Background/mechanistic context (islet tissue, not PBMC; no ancestry testing)
T2D islet scRNA-seq biomarker panel (Tang 2026)	Tang et al. identify PNLIP, BUB1, CTSB, NAMPT from islet scRNA-seq + LASSO, with qRT-PCR expression checks in a small peripheral blood/serum cohort. This provides islet-derived candidate genes with limited cross-compartment expression support, not PBMC immune validation.	Tang et al. 2026	single-study	Background/mechanistic context (islet tissue; blood/serum qRT-PCR is not PBMC immune profiling or ancestry validation)
Ancestry-specific T2D genetic mechanisms (same cohort)	Our prior work (Markelova et al.) applied Smith et al.’s pPGS to the same Chechen/Tatar/Yakut cohort used in the current project and found ancestry-associated distributions of T2D genetic clusters — Yakuts show higher beta-cell/lipid-related pPGS and lower obesity pPGS, while Chechens and Tatars show relatively more obesity/insulin-resistance pPGS patterns. These pPGS data are directly available for within-person correlation with PBMC immune features.	Markelova et al. 2025	own prior work, same cohort	Ancestry-motivation framing — provides same-cohort genetic context for testing mechanism-linked PBMC associations; does not by itself demonstrate ancestry-specific PBMC immune biology
East Asian vs. European T2D pathophysiology	Yabe et al. review population-level evidence that T2D in East Asian cohorts often involves lower insulin secretory capacity and lower insulin resistance than Caucasian comparison cohorts. This framework provides the literature backdrop for why ancestry-aware T2D mechanism analysis matters, and it is consistent with the beta-cell/lipid vs. obesity/insulin-resistance pPGS hypotheses from same-cohort prior work (Markelova et al.).	Yabe et al. 2015	Review (comprehensive)	Introduction/motivation — supports population-level pathophysiology framing, not genetic causation or PBMC immune evidence

Draft Argument Skeleton

1. T2D has measurable immune correlates in accessible blood immune compartments.
2. PBMC assays can reveal both compositional and state-based immune changes, but mixed-cell measurements require careful interpretation.
3. Ancestry-associated differences may be biologically relevant, socially mediated, technically confounded, or a combination.
4. Strong claims require ancestry-aware design, transparent covariate control, and replication across cohorts.

Contradictions To Watch

• Different studies may report opposite directions for cell proportions or pathway activity.
• Bulk PBMC findings may reflect cell-composition shifts rather than per-cell regulation.
• Ancestry-associated differences may disappear after controlling for site, batch, BMI, glycemia, medication, or socioeconomic variables.
• Whole-blood neutrophil-enrichment signals may not map directly onto PBMC-only data because PBMCs exclude most mature neutrophils.
• Cross-study meta-analysis can reveal weak coordinated effects, but these discoveries may still depend on sample selection, covariate modeling, and harmonization assumptions.
• Li et al. 2025 reports increased monocytes in T2D, while Gu et al. 2024’s cell-composition summary emphasizes different monocyte subset directions; differences may reflect control selection, subset definitions, or reanalysis design.
• Tkachenko et al. 2024 reports NK depletion in T2D PBMCs, while Gu et al. 2024 found a different NK composition trend; discrepancy may reflect sample size (n=2/group vs. larger cohort), population (Russian vs. Korean), or technical factors.
• Reanalysis of the same T2D dataset can add mechanistic layers without adding independent disease-cohort replication.
• Small single-center PBMC scRNA-seq studies such as Zhao and Fang 2025 can recover pathway hypotheses, but their directionality may not replicate in larger or ancestry-aware cohorts.

Ingested Literature Anchors

• Gu et al. 2024 anchors the claim that single-cell PBMC profiling can detect T2D-associated inflammatory immune remodeling in monocytes, cytotoxic T cells, and B cells.
• Gu et al. should not be used as direct ancestry evidence because the study does not perform ancestry-stratified or multi-ancestry analysis.
• Tkachenko et al. 2025 anchors the claim that T2D blood transcriptomic signatures are heterogeneous across cohorts, while meta-analysis highlights immune and stress-response pathways relevant to T2D.
• Tkachenko et al. should not be used as direct ancestry evidence because it does not perform ancestry-stratified analysis.
• Tkachenko et al. 2024 anchors the claim that PBMC scRNA-seq composition findings can differ between studies even with similar methodology, reinforcing the need for replication in the paper’s own analyses.
• Tkachenko et al. 2024 should not be used as direct ancestry evidence; its Russian cohort is treated as a single group without ancestry stratification.
• Tkachenko et al. 2024 provides the primary data (GSE280401, GSE280402) from the same research group that published the meta-analysis in Tkachenko et al. 2025; the two papers are complementary rather than independent.
• Li et al. 2025 anchors an immunometabolic reanalysis of T2D PBMC single-cell data, including T-cell metabolic subtypes, CellChat-inferred T-cell-monocyte communication, inferred TF activity, and drug-enrichment hypotheses.
• Li et al. 2025 should not be used as direct ancestry evidence because it does not test ancestry-stratified effects and its T2D cases overlap GSE268210 from Gu et al. 2024.
• Huang et al. 2022 provides T2D biomarker candidates and islet-level transcriptome context; uses pancreatic tissue rather than PBMCs, so it complements rather than directly supports PBMC ancestry claims.
• Huang et al. 2022 should not be used as direct ancestry evidence because it does not test ancestry-stratified effects.
• Tang et al. 2026 provides a second islet-derived T2D biomarker panel (PNLIP, BUB1, CTSB, NAMPT) using an scRNA-seq-first + LASSO pipeline, with cross-tissue validation into blood via qRT-PCR.
• Tang et al. 2026 should not be used as direct ancestry evidence because it uses a single-ancestry (Chinese) cohort and does not test ancestry-stratified effects.
• Markelova et al. 2025 (our prior work) anchors the claim that pPGS-defined T2D mechanism profiles vary by ancestry group in the same cohort studied in the current PBMC project, providing directly integrable genetic context for ancestry-aware PBMC immune analysis.
• Markelova et al. 2025 itself is a genotyping-only study without PBMC or immune data. Its value for the current project is that it provides pre-existing genetic characterization of the same cohort, enabling within-person integration of pPGS with newly acquired PBMC multi-omic data.
• Yabe et al. 2015 anchors a population-level framework in which T2D pathophysiology often differs between East Asian and Caucasian comparison cohorts, providing the literature framework for ancestry-aware T2D mechanism analysis.
• Yabe et al. should not be used as direct PBMC or immune evidence — it does not address immune cell biology. Its value is in establishing the ancestry-specific T2D pathophysiology framework that motivates the project’s central question.
• Yabe et al.’s documentation of East Asian T2D physiology and genetic-loci context supports ancestry-aware hypothesis generation, but it does not establish genetic causation or PBMC relevance.
• Zhao and Fang 2025 anchors a small Chinese PBMC scRNA-seq study reporting T-cell and monocyte inflammatory pathway signals in T2DM, including TNF/NF-kB, interferon-gamma response, T-cell receptor signaling, chemokine signaling, and TNFRSF1A-centered T-cell network interactions.
• Zhao and Fang should not be used as direct ancestry evidence because the paper does not perform ancestry-stratified analysis or genetic ancestry inference.

Next Source Targets

• Multi-ancestry immune profiling studies.
• Additional T2D PBMC bulk RNA-seq or single-cell studies to test whether Gu et al. findings replicate.
• Cohort papers with explicit ancestry methods.
• Reviews on immunometabolism and T2D that can support the introduction.
• Islet transcriptome studies as complementary tissue context for T2D molecular pathology.
• Additional multi-ancestry pPGS studies across diverse populations — our prior work (Markelova et al.) shows the approach for Russian populations; broader multi-ancestry replication would strengthen the paper introduction’s generalizability framing. (For the same-cohort within-person analysis, our own data is sufficient; external studies support the broader principle.)