# Scientific Gdc Portal > NCI Genomic Data Commons ポータルスキル。GDC REST API を用いたがんゲノムプロジェクト横断検索・ケースメタデータ・ 体細胞変異 (SSM)・遺伝子発現・ファイル取得。 ToolUniverse 連携: gdc。 - Skill: `nahisaho/scientific-gdc-portal-2` (Agent Skill) - Install (CLI): `npx skillmds@latest add nahisaho/scientific-gdc-portal-2` - Raw SKILL.md: https://api.skillmd.com/api/skills/nahisaho/scientific-gdc-portal-2/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Integrations & APIs - Author: nahisaho (https://skillmd.com/u/nahisaho) - Updated: 2026-09-21 - Page: https://skillmd.com/skills/nahisaho/scientific-gdc-portal-2 --- # Scientific GDC Portal NCI Genomic Data Commons (GDC) REST API を活用した がんゲノムプロジェクト横断検索・ケースメタデータ取得・ 体細胞変異 (SSM)・遺伝子発現パイプラインを提供する。 ## When to Use - TCGA/TARGET 等のがんゲノムデータを横断検索するとき - がん種別のケースメタデータを取得するとき - 特定遺伝子の体細胞変異 (SSM) 頻度を調べるとき - がんプロジェクトの統計サマリーを取得するとき - GDC ファイルメタデータを検索してダウンロード URL を取得するとき --- ## Quick Start ## 1. プロジェクト検索・統計 ```python import requests import pandas as pd GDC_API = "https://api.gdc.cancer.gov" def gdc_projects(disease_type=None, limit=50): """ GDC — プロジェクト検索。 Parameters: disease_type: str — 疾患タイプフィルタ (例: "Breast Invasive Carcinoma") limit: int — 最大結果数 """ url = f"{GDC_API}/projects" params = { "size": limit, "fields": ("project_id,name,primary_site," "disease_type,summary.case_count," "summary.file_count"), } if disease_type: params["filters"] = ( '{"op":"=","content":{"field":' f'"disease_type","value":"{disease_type}"}}}}' ) resp = requests.get(url, params=params, timeout=30) resp.raise_for_status() data = resp.json() rows = [] for hit in data.get("data", {}).get("hits", []): summary = hit.get("summary", {}) rows.append({ "project_id": hit.get("project_id", ""), "name": hit.get("name", ""), "primary_site": "; ".join( hit.get("primary_site", [])), "disease_type": "; ".join( hit.get("disease_type", [])), "case_count": summary.get( "case_count", 0), "file_count": summary.get( "file_count", 0), }) df = pd.DataFrame(rows) print(f"GDC projects: {len(df)}") return df ``` ## 2. ケースメタデータ ```python def gdc_cases(project_id, limit=100): """ GDC — ケースメタデータ取得。 Parameters: project_id: str — プロジェクト ID (例: "TCGA-BRCA") limit: int — 最大結果数 """ url = f"{GDC_API}/cases" filters = { "op": "=", "content": { "field": "project.project_id", "value": project_id, }, } params = { "filters": str(filters).replace("'", '"'), "fields": ("case_id,submitter_id," "demographic.gender," "demographic.race," "demographic.vital_status," "diagnoses.primary_diagnosis," "diagnoses.tumor_stage," "diagnoses.age_at_diagnosis"), "size": limit, } resp = requests.get(url, params=params, timeout=30) resp.raise_for_status() data = resp.json() rows = [] for hit in data.get("data", {}).get("hits", []): demo = hit.get("demographic", {}) or {} diag = (hit.get("diagnoses", [{}]) or [{}])[0] rows.append({ "case_id": hit.get("case_id", ""), "submitter_id": hit.get("submitter_id", ""), "gender": demo.get("gender", ""), "race": demo.get("race", ""), "vital_status": demo.get( "vital_status", ""), "diagnosis": diag.get( "primary_diagnosis", ""), "stage": diag.get("tumor_stage", ""), "age_at_diagnosis": diag.get( "age_at_diagnosis", ""), }) df = pd.DataFrame(rows) print(f"GDC cases: {project_id} → {len(df)}") return df ``` ## 3. 体細胞変異 (SSM) 検索 ```python def gdc_ssm_by_gene(gene_symbol, project_id=None, limit=100): """ GDC — 遺伝子別体細胞変異検索。 Parameters: gene_symbol: str — 遺伝子シンボル (例: "TP53") project_id: str — プロジェクト ID フィルタ limit: int — 最大結果数 """ url = f"{GDC_API}/ssms" filters = { "op": "and", "content": [ { "op": "=", "content": { "field": "consequence.transcript." "gene.symbol", "value": gene_symbol, }, } ], } if project_id: filters["content"].append({ "op": "=", "content": { "field": "cases.project.project_id", "value": project_id, }, }) params = { "filters": str(filters).replace("'", '"'), "fields": ("ssm_id,genomic_dna_change," "consequence.transcript.aa_change," "consequence.transcript." "consequence_type," "consequence.transcript." "gene.symbol"), "size": limit, } resp = requests.get(url, params=params, timeout=30) resp.raise_for_status() data = resp.json() rows = [] for hit in data.get("data", {}).get("hits", []): for csq in hit.get("consequence", []): tx = csq.get("transcript", {}) rows.append({ "ssm_id": hit.get("ssm_id", ""), "genomic_change": hit.get( "genomic_dna_change", ""), "gene": tx.get("gene", {}).get( "symbol", ""), "aa_change": tx.get("aa_change", ""), "consequence_type": tx.get( "consequence_type", ""), }) df = pd.DataFrame(rows) print(f"GDC SSM: {gene_symbol} → {len(df)} variants") return df ``` ## 4. GDC 統合パイプライン ```python def gdc_pipeline(project_id, gene_symbol=None, output_dir="results"): """ GDC 統合パイプライン。 Parameters: project_id: str — プロジェクト ID gene_symbol: str — 遺伝子フィルタ output_dir: str — 出力ディレクトリ """ from pathlib import Path output_dir = Path(output_dir) output_dir.mkdir(parents=True, exist_ok=True) # 1) プロジェクト情報 projects = gdc_projects() projects.to_csv(output_dir / "gdc_projects.csv", index=False) # 2) ケースメタデータ cases = gdc_cases(project_id) cases.to_csv(output_dir / "gdc_cases.csv", index=False) # 3) 体細胞変異 if gene_symbol: ssm = gdc_ssm_by_gene(gene_symbol, project_id) ssm.to_csv(output_dir / "gdc_ssm.csv", index=False) print(f"GDC pipeline: {project_id} → {output_dir}") return {"cases": cases} ``` --- ## ToolUniverse 連携 | TU Key | ツール名 | 連携内容 | |--------|---------|---------| | `gdc` | GDC | NCI Genomic Data Commons REST API | ## パイプライン統合 ``` cancer-genomics → gdc-portal → precision-oncology (COSMIC/DepMap) (GDC API) (MTB レポート) │ │ ↓ icgc-cancer-data ───────┘ variant-interpretation (ICGC DCC) (ClinVar/ACMG) ``` ## パイプライン出力 | ファイル | 説明 | 次スキル | |---------|------|---------| | `results/gdc_projects.csv` | プロジェクト一覧 | → cancer-genomics | | `results/gdc_cases.csv` | ケースメタデータ | → precision-oncology | | `results/gdc_ssm.csv` | 体細胞変異 | → variant-interpretation |