Scrnaseq Pd1 Analysis

scRNA-seq PD-1 Immunotherapy Analysis

Automated pipeline for analyzing single-cell RNA sequencing data in immunotherapy response studies.

Trigger Detection

Proactively offer this workflow when detecting:

• GEO dataset ID (GSE format) + any of: scRNA-seq, single-cell, PD-1, immunotherapy, tumor, cancer

Example prompt: "检测到GEO单细胞数据集。是否运行完整分析流程？包括聚类、差异表达、通路富集、细胞相互作用及发表级报告。"

Analysis Pipeline

Step 1: Data Acquisition → Download GEO data, QC filtering, create Seurat object
Step 2: Clustering       → SCTransform, PCA, UMAP, Louvain clustering, annotation
Step 3: Differential Expr → Pseudobulk aggregation, Welch's t-test, FDR correction
Step 4: Pathway Analysis → Hypergeometric test on 21 immune pathways
Step 5: Biomarkers       → Checkpoint molecules, activation/exhaustion markers
Step 6: Interactions     → 36 ligand-receptor pairs, network visualization
Step 7: Report           → Methods, figure legends, PDF report, archive

Quick Start

1. Initialize project:

python scripts/init_project.py GSE243013 --output ./analysis

1. Run analysis (modify config as needed):

Rscript scripts/run_analysis.R --config config.yaml

1. Generate report:

python scripts/generate_report.py --input ./results --output ./report

Key Parameters

| Component | Parameter | Default | Description | |-----------|-----------|---------|-------------| | QC | pctcountsmt | <20% | Mitochondrial content threshold | | Clustering | resolution | 0.5 | Louvain resolution | | UMAP | n_neighbors | 30 | Neighbor count | | DEG | FDR | <0.05 | Significance threshold | | Pathway | FDR | <0.20 | Enrichment threshold |

For complete parameter documentation, see [references/parameters.md](references/parameters.md).

Output Structure

results/
├── figures/          # 28 publication-ready figures
├── tables/           # 20 analysis tables (DEG, pathways, interactions)
├── report/           # PDF report, methods, figure legends
└── archive/          # Logs, session info, parameters

For detailed output specifications, see [references/output-files.md](references/output-files.md).

Requirements

• R ≥4.3 with Seurat v5
• RAM: 64GB minimum, 476GB+ for 1M+ cells
• Storage: 100GB+ per dataset

Troubleshooting

| Issue | Solution | |-------|----------| | Memory error | Use CreateSketch() for large datasets | | pivot_wider error | Scripts use tidyr::spread() (fixed) | | Long runtime | 20-60 min per step for 1M+ cells |

References

• Seurat v5: Hao et al., Cell 2021
• Pseudobulk DEG: Squair et al., BioRxiv 2021
• SCTransform: Hafemeister & Satija, Genome Biol 2019

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: TingxuanGu
• Source: TingxuanGu/scRNA-seq-Analysis-Skill-for-Claude-Code
• License: MIT

Install and usage instructions live in the source repository linked above.