tcga-gene-expression - SKILL.md Agent Skill

name: TCGA Gene Expression description: Query and analyze tumor gene expression profiles from The Cancer Genome Atlas (TCGA). Supports cohort-level expression lookup, tumor-versus-normal comparison, subtype stratification, and candidate biomarker exploration across cancer types. metadata: scpToolId: tcga-gene-expression scpCategory: life_science scpType: database provider: 上海人工智能实验室

TCGA Gene Expression

Overview

Queries TCGA gene expression datasets for tumor cohorts and matched clinical contexts. The tool supports expression lookup across cancer types, tumor versus normal comparison, subtype-aware stratification, and biomarker-oriented interpretation of transcriptomic patterns.

MCP Invocation

# Python / FastMCP
from server.api.scp-tools.invoke import invokeMcpTool

result = await invokeMcpTool(
    toolId="tcga-gene-expression",
    action="query",
    params={
        "gene": "EGFR",                 # Gene symbol or Ensembl ID
        "cancer_type": "LUAD",          # TCGA cohort code
        "comparison": "tumor_vs_normal",# "tumor" | "normal" | "tumor_vs_normal"
        "summary_stat": "median",       # "median" | "mean" | "distribution"
        "include_survival": True,       # Optional clinical association
        "include_subtypes": True        # Stratify by molecular subtype if available
    }
)

Local Description

网页端调用: https://scphub.intern-ai.org.cn/skill/tcga-gene-expression

下载说明: 本地为 SKILL.md 资源定义，工具执行由 SCP Hub 远程服务提供。需配置 openteam.json → integrations.scpHub.apiKey。

Output

Cohort-level expression values for the queried gene
Tumor versus normal differential expression summaries
Cancer subtype stratification results when available
Survival or clinical association hints for biomarker analysis
Exportable expression statistics and cohort metadata