單細胞 RNA 定序會產生複雜資料集，需要專門的分析。此技能提供完整流程，涵蓋品質控制、降維、分群與單細胞基因表達資料的視覺化。
스킬 ZIP 다운로드
Claude에서 업로드
설정 → 기능 → 스킬 → 스킬 업로드로 이동
토글을 켜고 사용 시작
테스트해 보기
"scanpy" 사용 중입니다. Load my single-cell data and perform quality control
예상 결과:
- Loaded 3,245 cells x 20,000 genes
- QC metrics: mean 1,542 genes per cell, 4.2% mitochondrial reads
- After filtering: 2,987 cells x 15,432 genes (92% cells retained)
- Saved QC violin plots to figures/qc_violin.pdf
"scanpy" 사용 중입니다. Run complete clustering and annotation workflow
예상 결과:
- Identified 12 cell clusters using Leiden algorithm
- Generated UMAP visualization colored by cluster
- Top marker genes identified for each cluster
- Cell types annotated based on known marker expression
보안 감사
안전All 228 static findings are false positives. This is a legitimate scientific computing skill for single-cell RNA-seq analysis. The scanner incorrectly flagged: markdown inline code formatting (backticks), file I/O functions for data reading, directory creation operations, and git tree hashes as C2 indicators. No malicious patterns, network exfiltration, or command injection risks exist after human evaluation.
위험 요인
📁 파일 시스템 액세스 (2)
🌐 네트워크 접근 (1)
품질 점수
만들 수 있는 것
探索性 scRNA-seq 分析
分析單細胞基因表達資料集以辨識細胞類型、狀態與族群。
標記基因探索
找出群集之間的差異表達基因並刻畫細胞族群。
視覺化流程
產生用於發表的 UMAP、t-SNE 與其他降維圖。
이 프롬프트를 사용해 보세요
Load my single-cell data from data.h5ad and show me the basic structure including cell and gene counts.
Run quality control on my dataset, filter cells with less than 200 genes or more than 5% mitochondrial reads, and generate QC plots.
Perform clustering at resolution 0.5, generate UMAP visualization, and identify marker genes for each cluster.
Run a complete scanpy workflow: QC, normalization, highly variable genes, PCA, neighbors, UMAP, Leiden clustering at resolution 0.8, and save results.
모범 사례
- 在過濾前一定要保存原始計數：adata.raw = adata
- 透過檢查已知標記基因表達來驗證分群
- 保存中間結果以避免重跑耗時流程
피하기
- 在下游分析前跳過品質控制步驟
- 未測試多個值就使用預設分群解析度
- 未在過濾前後進行資料視覺化