scRNA-seq Tumor Microenvironment Results

Task Info

Task ID: writing-scrna-biology-en-l3
Discipline: Biology / Cancer Immunology
Language: English
Section Type: Results Paragraph
Verdict: PASS
Test Depth: L3
Test Date: 2026-06-15

Summary

English Results paragraph on scRNA-seq tumor microenvironment analysis. 67,842 cells, 15 clusters, p=0.003 Wilcoxon test, Spearman rho=0.67 TAM-exhaustion correlation.

Generated Output

# Output: scRNA-seq Tumor Microenvironment — Results Paragraph

Visual Preview

(Text-only writing task, no visual artifacts)

Results Paragraph (English academic writing)

To characterize the cellular composition and transcriptional heterogeneity of the tumor microenvironment, we performed single-cell RNA sequencing (scRNA-seq) on freshly dissociated tumor samples from 8 treatment-naive patients using the 10x Genomics Chromium platform. After quality control filtering (nGenes >200, <5000; mitochondrial content <20%), a total of 67,842 high-quality cells were retained for downstream analysis (Supplementary Fig. 1a-b).

Unsupervised clustering using the Louvain algorithm identified 15 distinct cell clusters (Fig. 1a), which were annotated based on canonical marker gene expression (Fig. 1b). The major cell types included CD8+ T cells (18.3% of total cells), CD4+ T cells (12.7%), regulatory T cells (Tregs, 4.2%), B cells (6.8%), natural killer (NK) cells (3.1%), tumor-associated macrophages (TAMs, 15.4%), dendritic cells (DCs, 2.3%), cancer-associated fibroblasts (CAFs, 8.9%), and epithelial tumor cells (28.3%) (Fig. 1c). Notably, the proportion of exhausted CD8+ T cells, defined by co-expression of PDCD1, LAG3, and TIGIT, was significantly higher in advanced-stage (III-IV) tumors compared to early-stage (I-II) tumors (mean 23.1% vs. 8.7%, p=0.003, Wilcoxon rank-sum test; Fig. 1d-e).

TAMs exhibited a continuous spectrum of polarization states rather than discrete M1/M2 categories (Fig. 2a). Pseudotime trajectory analysis revealed three major TAM trajectories: an inflammatory (M1-like) branch enriched for IL1B, TNF, and CXCL10; an immunosuppressive (M2-like) branch characterized by expression of MRC1, CD163, and IL10; and an intermediate state that co-expressed markers of both programs (Fig. 2b-c). The immunosuppressive TAM branch was positively correlated with CD8+ T cell exhaustion scores (Spearman rho=0.67, p<0.001; Fig. 2d), suggesting a potential functional coupling between TAM polarization and T cell dysfunction in the tumor microenvironment.

This is a text-only writing skill. No visual artifacts are produced.