TCGA 与 ICGC 数据联合挖掘¶

# === 使用 TCGAbiolinks 下载 TCGA 数据 ===
library(TCGAbiolinks)                  # TCGA数据访问包

# 1. 查询可用项目
projects <- getGDCprojects()           # 获取所有项目
tcga_projects <- projects[grep("TCGA", projects$id), ]  # 筛选TCGA

# 2. 下载表达数据（以肝癌为例）
query_expr <- GDCquery(
  project = "TCGA-LIHC",              # 肝细胞癌
  data.category = "Transcriptome Profiling",  # 转录组
  data.type = "Gene Expression Quantification",  # 基因表达
  workflow.type = "STAR - Counts"      # STAR计数
)
GDCdownload(query_expr)               # 下载数据
expr_data <- GDCprepare(query_expr)    # 整理为 SummarizedExperiment

# 3. 下载突变数据
query_mut <- GDCquery(
  project = "TCGA-LIHC",              # 肝细胞癌
  data.category = "Simple Nucleotide Variation",
  data.type = "Masked Somatic Mutation"
)
GDCdownload(query_mut)
maf_data <- GDCprepare(query_mut)      # MAF数据

# 4. 下载临床数据
clinical <- GDCquery_clinic("TCGA-LIHC", type = "clinical")

# === ICGC 数据下载与整合 ===
import pandas as pd                    # 数据处理
import requests                        # API请求

# 1. ICGC DCC API 查询
base_url = "https://dcc.icgc.org/api/v1"

# 查询肝癌项目（LIRI-JP: 日本肝癌）
resp = requests.get(f"{base_url}/projects/LIRI-JP")
project_info = resp.json()             # 获取项目信息
print(f"项目: {project_info['name']}")
print(f"样本数: {project_info['totalDonorCount']}")

# 2. 下载简单体细胞突变
mutations_url = (f"{base_url}/download?fn=/release_28/"
                 f"Projects/LIRI-JP/simple_somatic_mutation.open.LIRI-JP.tsv.gz")
# 注意：受控数据需要 DACO 访问权限

# 3. 统一格式处理（ICGC → 类MAF格式）
def icgc_to_maf_format(icgc_df):
    """将 ICGC 突变格式转为类 MAF 格式"""
    maf = pd.DataFrame({
        "Hugo_Symbol": icgc_df["gene_affected"],       # 基因名
        "Chromosome": icgc_df["chromosome"],            # 染色体
        "Start_Position": icgc_df["chromosome_start"],  # 起始位置
        "End_Position": icgc_df["chromosome_end"],      # 终止位置
        "Reference_Allele": icgc_df["reference_genome_allele"],  # 参考等位
        "Tumor_Seq_Allele2": icgc_df["mutated_to_allele"],       # 突变等位
        "Variant_Classification": icgc_df["consequence_type"],    # 变异类型
        "Tumor_Sample_Barcode": icgc_df["icgc_donor_id"],        # 样本ID
    })
    return maf

# 4. 联合分析：TCGA + ICGC 突变频率比较
def compare_mutation_freq(tcga_maf, icgc_maf, top_n=20):
    """比较两个队列的基因突变频率"""
    # TCGA 突变频率
    tcga_samples = tcga_maf["Tumor_Sample_Barcode"].nunique()
    tcga_freq = (tcga_maf.groupby("Hugo_Symbol")
                 ["Tumor_Sample_Barcode"].nunique() / tcga_samples)

    # ICGC 突变频率
    icgc_samples = icgc_maf["Tumor_Sample_Barcode"].nunique()
    icgc_freq = (icgc_maf.groupby("Hugo_Symbol")
                 ["Tumor_Sample_Barcode"].nunique() / icgc_samples)

    # 合并比较
    comparison = pd.DataFrame({
        "TCGA_freq": tcga_freq,        # TCGA频率
        "ICGC_freq": icgc_freq         # ICGC频率
    }).fillna(0)                       # 缺失填0

    # 排序并打印
    comparison["mean_freq"] = (comparison["TCGA_freq"] +
                                comparison["ICGC_freq"]) / 2
    comparison = comparison.sort_values("mean_freq", ascending=False)
    print(f"\n{'基因':<15} {'TCGA':>8} {'ICGC':>8}")
    print("-" * 35)
    for gene, row in comparison.head(top_n).iterrows():
        print(f"{gene:<15} {row['TCGA_freq']:>7.1%} {row['ICGC_freq']:>7.1%}")
    return comparison

# compare_mutation_freq(tcga_maf, icgc_maf)