Systematic mapping of emergent transcriptional states in interacting single-cell dyads by Cell-Cell-seq
About this tab ▼
UMAP of 9,963 nanovial cells from Cell-Cell-seq Run 2. Left panel displays UMAP by sample populations.
Right panel overlays gene module scores or gene-pair co-expression (per-cluster Pearson r).
Toggle Leiden Clusters to see cluster boundaries. Click population chips to isolate.
Populations
Gene Signatures
Clusters shaded by Pearson r
LowHigh
Cell Info drag to move
Up-regulated
Down-regulated
NS
Min |log₂FC|≥ 1.0
Min −log₁₀(p)≥ 10
Labels:
Transcriptome-wide differential gene expression comparing interacting T cell–PC3 tumor dyads
vs. resting pseudo-dyads captured by Cell-Cell-seq, using Seurat Wilcoxon rank-sum test
(FindMarkers), summarized over 50 independent differential expression tests.
Each volcano shows ~5,000 genes per cell type.
▲ Up-regulated genes are enriched in true interacting dyads;
▼ down-regulated genes are relatively suppressed compared to pseudo-dyads.
Key T cell findings: immediate-early transcription factors (EGR1/3, IER3) and chemokines
(CCL3/4/5, XCL1/2) are upregulated — a unique activated state captured only by Cell-Cell-seq.
Use category filters above to highlight genes by biological role.
T Cell
PC3 Cell
◢ Click a gene or drag to box-select — results appear here
About this tab ▼
Compares early activation gene induction in Cell-Cell-seq (dyad vs pseudo-dyad) versus
standard co-culture (co-cultured vs unactivated T cells). The bar chart shows mean normalized
expression of six immediate-early genes, with fold-change between conditions. The ranked plots below
visualize per-cell expression distributions of EGR1 and DUSP2 across four population groups,
illustrating how Cell-Cell-seq captures stronger activation signatures than standard co-culture.
View
Dataset
Gene Groups
T cell gene
PC3 cell gene
Ribbon
Positive r
Negative r
Width = |r| strength
Min |correlation|
≥ 0.14
Dataset
Color Scale
−1
+1
Pearson correlations in true interacting dyads (Fig. 5E).
About ▼
Gene–gene Pearson correlations computed across thousands of tumor–T cell dyads captured by Cell-Cell-seq (Fig. 5E).
Key finding: positive correlations between T cell chemokines (CCL3/4, XCL1/2) and
tumor chemokines (CXCL1/3/8) indicate bidirectional paracrine signaling during active immune engagement.
Negative correlations between tumor resistance gene TM4SF1 and T cell activation markers suggest
active suppression of T cell function.
About this tab ▼
Cells ordered along a continuum of relative cell-type signature strength, adapted from Baghdasarian et al. 2026 Fig. 3D (Run 1) and recomputed here on Run 2 nanovial data (9,963 cells, Leiden res=0.5).
Four tracks show per-cell module scores normalized so they sum to 1 per cell:
PC3 (green), T cell (blue), Cytokine (red), IRF (amber).
Hover waterfall sections or UMAP cluster labels to cross-highlight.
Module scores computed by Seurat AddModuleScore() on normalized expression data.
UMAP
Cell Continuum — ordered by PC3 → T cell balance
Contact-Dependent Signaling Genes
Three-way detection comparison — click a region or category to filter
CellChat (activated T vs PC3): CellChat run on activated T cells and PC3 cells. CellChat (standard co-culture): CellChat run on standard co-culture populations. Pseudo/real (Cell-Cell-seq): emergent genes from differential expression of interacting dyads versus resting pseudo-dyads. 344 pseudo/real-only genes represent transcriptional states that emerge specifically from cell-cell contact — a key finding of this study.
T = in co-expression panel
PC3 = in co-expression panel
ℹ How to use this panel
Venn regions — click a numbered area on the diagram or a button below it to filter the gene list to that overlap. Click again to reset to all genes.
Category pills — click a colored pill to focus on one gene family. Venn counts update live to reflect only genes in that category. Click again to clear.
Gene chips — click any chip to see a functional annotation and a GeneCards link. Bold chips with T / PC3 badges are genes also featured in the co-expression analysis.
Search — type ≥2 characters to highlight matching genes across all categories.
All Detected Signaling Genes
Grouped by category — click a Venn region or category pill to filter