Figure Plan

Working plan for main figures and graphical abstract.
Figures 3–4 are provisional and may be reorganized as results solidify.


Graphical Abstract

Concise visual summary of the study: ancestry-aware multi-modal PBMC profiling → pan-ancestry + ancestry-specific immune signals.


Figure 1 — Cohort & Study Design

  • (a) Map of the Russian Federation with sampling sites marked.
  • (b) PCA of genotyping data colored by genetic ancestry cluster (linking to Markelova et al. 2025); potentially annotated with metadata (sex, cohort site).
  • (c) scRNA-seq UMAP colored by major cell type clusters.
  • (d) Experimental schema: PBMCs → scRNA-seq + genotyping → downstream analyses.

todo add pooling —> demultiplexing


Figure 2 — Pan-Ancestry T2D Effects (Convergent Signal)

Common T2D-associated immune changes reproducible across all ancestry groups.

  • (a) UMAP from MiloPy colored by logFC (common T2D abundance shifts).
  • (b) Boxplots of major cell-type proportion changes across ancestries.
  • (c) Dotplot of pathways with significant convergent changes (model: ~ T2D × ancestry + age + sex).
  • Extended analyses (subset to convergent results): Progeny pathway activity, TF SCENIC regulons, pseudotime clusters, cell–cell interaction strength.

Figure 3 — Ancestry-Specific Gain-of-Function Mechanisms

Immune features that are enhanced or uniquely present in specific ancestry groups.

  • Differential abundance (MiloPy) — expansions unique to one ancestry.
  • Pathway enrichment — processes upregulated in an ancestry-specific manner.
  • TF SCENIC regulons active only in particular groups.
  • Cell–cell interaction strength gained in specific ancestries.

Figure 4 — Ancestry-Specific Loss/Attenuation Mechanisms

Immune features that are diminished, absent, or failed regulatory mechanisms in specific ancestry groups.

  • Differential abundance — depleted cell populations.
  • Pathway enrichment — processes that fail to activate or are suppressed.
  • TF SCENIC regulons missing or attenuated in specific ancestries.
  • Pseudotime trajectories truncated or shifted in particular groups.

Notes

  • Supplementary figures TBD (likely QC, marker validation, model diagnostics, full pathway/regulon tables).
  • Panel structure for Figures 3–4 is preliminary; organization may shift to population-based or analysis-method-based framing depending on which ancestry-specific findings are strongest.