{"id":61149,"date":"2024-10-17T11:39:33","date_gmt":"2024-10-17T03:39:33","guid":{"rendered":"http:\/\/www.biocloudservice.com\/wordpress\/?p=61149"},"modified":"2024-10-17T11:39:33","modified_gmt":"2024-10-17T03:39:33","slug":"%e7%94%9f%e4%bf%a1%e5%88%86%e6%9e%90%e5%bf%85%e5%a4%87%e9%94%a6%e5%9b%8a%ef%bc%81%e5%a4%a7%e6%b5%b7%e5%93%a5%e5%b8%a6%e4%bd%a0%e4%b8%80%e6%96%87get%e7%94%9f%e5%ad%98%e5%88%86%e6%9e%90%e5%bb%ba","status":"publish","type":"post","link":"http:\/\/www.biocloudservice.com\/wordpress\/?p=61149","title":{"rendered":"\u751f\u4fe1\u5206\u6790\u5fc5\u5907\u9526\u56ca\uff01\u5927\u6d77\u54e5\u5e26\u4f60\u4e00\u6587get\u751f\u5b58\u5206\u6790\u5efa\u6a21\uff01"},"content":{"rendered":"\n

\u524d\u8a00<\/strong><\/strong><\/p>\n\n\n\n

\u751f\u5b58\u5206\u6790\u6587\u7ae0\u4e2d\u6bd4\u6bd4\u7686\u662f\uff0c\u4f60\u7684\u5efa\u6a21\u65b9\u6cd5\u7528\u5bf9\u4e86\u5417\uff1f\u4f60\u7684\u6a21\u578b\u771f\u7684\u51c6\u786e\u5417\uff1f\u662f\u4e0d\u662f\u4e1c\u62fc\u897f\u51d1\u627e\u6765\u4ee3\u7801\uff0c\u53ef\u89c6\u5316\u4e4b\u540e\u8349\u8349\u4e86\u4e8b\uff1f\u5927\u6d77\u54e5\u6700\u7ecf\u542c\u5230\u5927\u5bb6\u8bf4\u751f\u5b58\u5206\u6790\u5efa\u6a21\u5b9e\u5728\u662f\u592a\u590d\u6742\u4e86\uff0c\u5bf9\u4e8e\u5efa\u6a21\u4e86\u89e3\u7684\u592a\u80a4\u6d45\uff0c\u4e1c\u62fc\u897f\u51d1\u624d\u80fd\u5b8c\u6210\u751f\u5b58\u5206\u6790\u5168\u90e8\u6d41\u7a0b\uff01\u770b\u5230\u5927\u5bb6\u7684\u547c\u58f0\uff0c\u5927\u6d77\u54e5\u8d76\u7d27\u653e\u4e0b\u624b\u91cc\u7684\u5b9e\u9a8c\uff0c\u8fde\u591c\u7ed9\u5927\u5bb6\u6574\u7406\u4e86\u4e00\u4efd\u751f\u5b58\u5206\u6790\u9526\u56ca\u3002\u4ece\u6570\u636e\u51c6\u5907\u5230\u6700\u7ec8\u6821\u51c6\u6a21\u578b\u7684\u5b8c\u6574\u5de5\u4f5c\u6d41\u7a0b\u5927\u6d77\u54e5\u90fd\u7ed9\u5927\u5bb6\u51c6\u5907\u597d\u4e86\uff01\u5bf9\u4e8e\u8bb8\u591a\u75be\u75c5\u6765\u8bf4\uff0c\u76ee\u6807\u4e8b\u4ef6\u7684\u53d1\u751f\u65f6\u95f4\uff08\u4f8b\u5982\u6b7b\u4ea1\u65f6\u95f4\u6216\u75be\u75c5\u8fdb\u5c55\uff09\u662f\u4e00\u4e2a\u5e7f\u6cdb\u4f7f\u7528\u7684\u7ec8\u70b9\u548c\u60a3\u8005\u9884\u540e\uff0c\u56e0\u6b64\uff0c\u8bc6\u522b\u57fa\u56e0\u7ec4\u3001\u5206\u5b50\u548c\u4e34\u5e8a\u6807\u5fd7\u7269\u4ee5\u9884\u6d4b\u60a3\u6709\u764c\u75c7\u7b49\u590d\u6742\u75be\u75c5\u7684\u60a3\u8005\u7684\u751f\u5b58\u6216\u8fdb\u5c55\u5df2\u7ecf\u53d8\u5f97\u5f88\u6d41\u884c\u3002\u5927\u6d77\u54e5\u7ed9\u5927\u5bb6\u63d0\u4f9b\u4e86\u4e00\u4efd\u9526\u56ca\u7528\u4e8e\u4f7f\u7528\u7ec4\u5b66\u548c\u6807\u51c6\u4e34\u5e8a\u6570\u636e\u8fdb\u884c\u751f\u5b58\u5206\u6790\uff0c\u7279\u522b\u5173\u6ce8\u751f\u5b58\u76f8\u5173\u7ec4\u5b66\u7279\u5f81\u7684\u7279\u5f81\u9009\u62e9\u548c\u751f\u5b58\u6a21\u578b\u9a8c\u8bc1\uff0c\u8fd8\u6db5\u76d6\u4e86\u8bb8\u591a\u7528\u4e8e\u7279\u5f81\u9009\u62e9\u548c\u751f\u5b58\u9884\u6d4b\u7684\u60e9\u7f5a\u56de\u5f52\u548c\u8d1d\u53f6\u65af\u6a21\u578b\uff0c\u5e76\u8003\u8651\u4e86\u5b83\u4eec\u7684\u5177\u4f53\u5047\u8bbe\u548c\u5e94\u7528\u3002\u5927\u6d77\u54e5\u5c31\u662f\u8fd9\u4e48\u7684\u5ba0\u7c89\uff0c\u6709\u4ec0\u4e48\u751f\u4fe1\u5206\u6790\u4e0a\u7684\u95ee\u9898\u5927\u5bb6\u5c3d\u7ba1\u54a8\u8be2\u5927\u6d77\u54e5\uff01\u6ca1\u6709\u65f6\u95f4\u5b66\u4e60\u7684\u5c0f\u4f19\u4f34\u4eec\u4e5f\u4e0d\u8981\u7740\u6025\u54e6\uff01\u6709\u9700\u8981\u751f\u4fe1\u5206\u6790\u7684\u5c0f\u4f19\u4f34\u4eec\u4e5f\u53ef\u4ee5\u627e\u5927\u6d77\u54e5\u54e6\uff01\u7ec3\u4e86\u5341\u5e74\u751f\u4fe1\u5206\u6790\u7684\u5927\u6d77\u54e5\u5bf9\u4e8e\u751f\u4fe1\u5206\u6790\u77e5\u8bc6\u5df2\u7ecf\u5982\u9c7c\u5f97\u6c34\u4ece\u5206\u6790\u5230\u53ef\u89c6\u5316\u76f4\u5230\u4f60\u6ee1\u610f\u4e3a\u6b62\uff01<\/p>\n\n\n\n

\"\"<\/figure>\n\n\n\n

\u751f\u4fe1\u6570\u636e\u5904\u7406\u8d77\u6765\u5360\u7528\u5185\u5b58\u5b9e\u5728\u592a\u5927\u4e86\uff0c\u653e\u8fc7\u81ea\u5df1\u7684\u7535\u8111\u5427\uff01\u5927\u6d77\u54e5\u5728\u8fd9\u91cc\u7ed9\u5927\u5bb6\u9001\u4e0a\u798f\u5229\u4e86\uff0c\u6709\u9700\u8981\u670d\u52a1\u5668\u7684\u5c0f\u4f19\u4f34\u4eec\uff0c\u6b22\u8fce\u5927\u5bb6\u8054\u7cfb\u5927\u6d77\u54e5\uff0c\u4fdd\u8bc1\u670d\u52a1\u5668\u7684\u6027\u4ef7\u6bd4\u6700\u9ad8\u54e6\uff01<\/p>\n\n\n\n

\u4ee3\u7801\u6559\u7a0b<\/strong><\/strong><\/p>\n\n\n\n

\u5c0f\u4f19\u4f34\u4eec\u6ce8\u610f\u5566\uff01\u7531\u4e8e\u5206\u4eab\u7bc7\u5e45\u6709\u9650\uff0c\u66f4\u8be6\u7ec6\u7684\u5efa\u6a21\u6d41\u7a0b\u4ee5\u53ca\u4ee3\u7801\u79c1\u4fe1\u5927\u6d77\u54e5\u83b7\u53d6\u54e6\uff01<\/p>\n\n\n\n

\u6211\u4eec\u5c06\u4f7f\u7528TCGA\u6570\u636e\u6765\u8bc1\u660e\uff1a<\/p>\n\n\n\n

    \n
  • \u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b<\/li>\n\n\n\n
  • \u751f\u5b58\u548c\u7ec4\u5b66\u6570\u636e\u7684\u9884\u5904\u7406<\/li>\n\n\n\n
  • \u901a\u8fc7\u7ecf\u5178\u7edf\u8ba1\u65b9\u6cd5\u5206\u6790\u751f\u5b58\u6570\u636e<\/li>\n\n\n\n
  • \u7ec4\u5b66\u6570\u636e\u7684\u65e0\u76d1\u7763\u5b66\u4e60<\/li>\n\n\n\n
  • \u751f\u5b58\u548c\u7ec4\u5b66\u6570\u636e\u7684\u9891\u7387\u548c\u8d1d\u53f6\u65af\u76d1\u7763\u5b66\u4e60<\/li>\n<\/ul>\n\n\n\n

    \u9996\u5148\uff0c\u6211\u4eec\u52a0\u8f7d\u672c\u6559\u7a0b\u4e2d\u4f7f\u7528\u7684\u6240\u6709\u5fc5\u8981\u5e93\u3002\u7136\u540e\uff0c\u6211\u4eec\u4f7f\u7528TCGAbiolinks\u8f6f\u4ef6\u5305\u4e2d\u7684\u51fd\u6570\u67e5\u8be2\u548c\u4e0b\u8f7d\u6765\u81ea\u591a\u79cd\u764c\u75c7\u7c7b\u578b\u7684TCGA\u751f\u5b58\u548c\u4e34\u5e8a\u6570\u636e\u3002\u8f93\u5165\u6570\u636e\u76f4\u63a5\u901a\u8fc7R\u5305TCGAbiolinks\u4e0b\u8f7d\u54e6\uff01<\/p>\n\n\n\n

    1.TCGA\u4e34\u5e8a\u548c\u751f\u5b58\u6570\u636e<\/strong><\/strong><\/p>\n\n\n\n

    #\u52a0\u8f7d\u6240\u6709\u7684R\u5305<\/p>\n\n\n\n

    library(“TCGAbiolinks”)        <\/p>\n\n\n\n

    library(“SummarizedExperiment”)<\/p>\n\n\n\n

    library(“DESeq2”)              <\/p>\n\n\n\n

    library(“M3C”)                 <\/p>\n\n\n\n

    library(“dplyr”)<\/p>\n\n\n\n

    library(“ggplot2”)<\/p>\n\n\n\n

    library(“survival”)<\/p>\n\n\n\n

    library(“survminer”)<\/p>\n\n\n\n

    library(“glmnet”)<\/p>\n\n\n\n

    library(“plotmo”)<\/p>\n\n\n\n

    library(“grpreg”)<\/p>\n\n\n\n

    library(“SGL”)<\/p>\n\n\n\n

    library(“psbcGroup”)<\/p>\n\n\n\n

    library(“psbcSpeedUp”)<\/p>\n\n\n\n

    library(“BhGLM”)<\/p>\n\n\n\n

    library(“risksetROC”)<\/p>\n\n\n\n

    library(“riskRegression”)<\/p>\n\n\n\n

    library(“peperr”)<\/p>\n\n\n\n

    library(“c060”)<\/p>\n\n\n\n

    library(“rms”)<\/p>\n\n\n\n

    library(“survAUC”)<\/p>\n\n\n\n

    library(“regplot”)<\/p>\n\n\n\n

    #\u4e0b\u8f7d\u4e34\u5e8a\u6570\u636e\uff0c\u4f7f\u7528GDC api\u65b9\u6cd5\u63d0\u53d6\u591a\u79cd\u80bf\u7624\u7684\u6570\u636e<\/p>\n\n\n\n

    cancer_types <- c(<\/p>\n\n\n\n

      “TCGA-BLCA”, “TCGA-BRCA”, “TCGA-COAD”, “TCGA-LIHC”,<\/p>\n\n\n\n

      “TCGA-LUAD”, “TCGA-PAAD”, “TCGA-PRAD”, “TCGA-THCA”<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    clin <- NULL<\/p>\n\n\n\n

    for (i in seq_along(cancer_types)) {<\/p>\n\n\n\n

      tmp <- TCGAbiolinks::GDCquery_clinic(project = cancer_types[i], type = “clinical”)<\/p>\n\n\n\n

      clin <- rbind(clin, tmp[, c(<\/p>\n\n\n\n

        “project”, “submitter_id”, “vital_status”,<\/p>\n\n\n\n

        “days_to_last_follow_up”, “days_to_death”,<\/p>\n\n\n\n

        “age_at_diagnosis”, “gender”, “race”,<\/p>\n\n\n\n

        “ethnicity”, “ajcc_pathologic_t”<\/p>\n\n\n\n

      )])<\/p>\n\n\n\n

    }<\/p>\n\n\n\n

    #\u63d0\u53d6\u6bcf\u4f4d\u60a3\u8005\u7684\u89c2\u5bdf\u65f6\u95f4\uff0c\u4ee5\u5e74\u4e3a\u5355\u4f4d<\/p>\n\n\n\n

    clin$time <- apply(clin[, c(“days_to_death”, “days_to_last_follow_up”)], 1, max, na.rm = TRUE) \/ 365.25<\/p>\n\n\n\n

    clin$age <- clin$age_at_diagnosis \/ 365.25<\/p>\n\n\n\n

    clin$status <- clin$vital_status<\/p>\n\n\n\n

    clin <- clin[, c(“project”, “submitter_id”, “status”, “time”, “gender”, “age”, “race”, “ethnicity”)]<\/p>\n\n\n\n

    # extract patients with positive overall survival time<\/p>\n\n\n\n

    clin <- clin[(clin$time > 0) & (clin$status %in% c(“Alive”, “Dead”)), ]<\/p>\n\n\n\n

    # \u60a3\u8005\u72b6\u51b5\u3001\u6027\u522b\u548c\u79cd\u65cf\u7684\u9891\u7387\u8868<\/p>\n\n\n\n

    clin %>%<\/p>\n\n\n\n

      dplyr::count(status, gender, ethnicity) %>%<\/p>\n\n\n\n

      group_by(status) %>%<\/p>\n\n\n\n

      mutate(prop = prop.table(n))<\/p>\n\n\n\n

    # \u6309\u764c\u75c7\u7c7b\u578b\u7b5b\u9009<\/p>\n\n\n\n

    ID <- 1:nrow(clin)<\/p>\n\n\n\n

    clin %>%<\/p>\n\n\n\n

      ggplot(<\/p>\n\n\n\n

        aes(y = ID, x = time, colour = project, shape = factor(status))<\/p>\n\n\n\n

      ) +<\/p>\n\n\n\n

      geom_segment(aes(x = time, y = ID, xend = 0, yend = ID)) +<\/p>\n\n\n\n

      geom_point() +<\/p>\n\n\n\n

      ggtitle(“”) +<\/p>\n\n\n\n

      labs(x = “Years”, y = “Patients”) +<\/p>\n\n\n\n

      scale_shape_discrete(name = “Status”, labels = c(“Censored”, “Dead”)) +<\/p>\n\n\n\n

      scale_color_discrete(<\/p>\n\n\n\n

        name = “Cancer”,<\/p>\n\n\n\n

        labels = c(“Bladder”, “Breast”, “Colon”, “Liver”, “Lung adeno”, “Pancreatic”, “Prostate”, “Thyroid”)<\/p>\n\n\n\n

      ) +<\/p>\n\n\n\n

      theme(legend.position = “top”, legend.direction = “vertical”) +<\/p>\n\n\n\n

      guides(color = guide_legend(nrow = 2, byrow = TRUE))<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    TCGA\u6cdb\u764c\u60a3\u8005\u7684\u603b\u751f\u5b58\u65f6\u95f4\u548c\u72b6\u51b5<\/p>\n\n\n\n

    \u6211\u4eec\u4f7f\u7528\u51fd\u6570\u6765\u67e5\u8be2\u548c\u4f7f\u7528\u4ee5\u53ca\u4e0b\u8f7d\u4e00\u79cd\u764c\u75c7\u7c7b\u578b\uff08\u4e73\u817a\u764c\uff09\u7684TCGA\u7ec4\u5b66\u6570\u636e\u3002\u51fd\u6570\u4e2d\u7684\u53c2\u6570\u6307\u5b9a\u7ec4\u5b66\u6570\u636e\u7684\u7c7b\u578b\u3002\u8bf7\u6ce8\u610f\uff0c\u4e0b\u8f7d\u7684\u7ec4\u5b66\u6570\u636e\u4f34\u968f\u7740\u5143\u6570\u636e\uff0c\u5305\u62ec\u751f\u5b58\u7ed3\u679c\u3001\u4e34\u5e8a\u548c\u4eba\u53e3\u7edf\u8ba1\u5b66\u53d8\u91cf\u3002<\/p>\n\n\n\n

    # \u4f7f\u7528GDC api\u65b9\u6cd5\u4e0b\u8f7dTCGA\u4e73\u817a\u764c(BRCA) mRNA-Seq\u6570\u636e<\/p>\n\n\n\n

    query <- TCGAbiolinks::GDCquery(<\/p>\n\n\n\n

      project = “TCGA-BRCA”,<\/p>\n\n\n\n

      data.category = “Transcriptome Profiling”,<\/p>\n\n\n\n

      data.type = “Gene Expression Quantification”,<\/p>\n\n\n\n

      workflow.type = “STAR – Counts”,<\/p>\n\n\n\n

      experimental.strategy = “RNA-Seq”,<\/p>\n\n\n\n

      sample.type = c(“Primary Tumor”)<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    TCGAbiolinks::GDCdownload(query = query, method = “api”)<\/p>\n\n\n\n

    dat <- TCGAbiolinks::GDCprepare(query = query)<\/p>\n\n\n\n

    SummarizedExperiment::assays(dat)$unstranded[1:5, 1:2]<\/p>\n\n\n\n

    \u5efa\u8bae\u5148\u5bf9RNA-seq\u6570\u636e\u91c7\u7528DESeq2\u6216TMM\u5f52\u4e00\u5316\u65b9\u6cd5\uff0c\u518d\u8fdb\u884c\u8fdb\u4e00\u6b65\u7edf\u8ba1\u5206\u6790\u3002\u5728\u8fd9\u91cc\uff0c\u6211\u4eec\u6f14\u793a\u4e86\u5982\u4f55\u4f7f\u7528R\/Bioconductor\u8f6f\u4ef6\u5305DESeq2\u6765\u89c4\u8303\u5316RNA\u8ba1\u6570\u6570\u636e\u3002<\/p>\n\n\n\n

    meta <- colData(dat)[, c(“project_id”, “submitter_id”, “age_at_diagnosis”, “ethnicity”, “gender”, “days_to_death”, “days_to_last_follow_up”, “vital_status”, “paper_BRCA_Subtype_PAM50”, “treatments”)]<\/p>\n\n\n\n

    meta$treatments <- unlist(lapply(meta$treatments, function(xx) {<\/p>\n\n\n\n

      any(xx$treatment_or_therapy == “yes”)<\/p>\n\n\n\n

    }))<\/p>\n\n\n\n

    dds <- DESeq2::DESeqDataSetFromMatrix(assays(dat)$unstranded, colData = meta, design = ~1)<\/p>\n\n\n\n

    dds2 <- DESeq2::estimateSizeFactors(dds)<\/p>\n\n\n\n

    RNA_count <- DESeq2::counts(dds2, normalized = TRUE)<\/p>\n\n\n\n

    RNA_count[1:5, 1:2]<\/p>\n\n\n\n

    \u4e3a\u4e86\u540c\u65f6\u4f7f\u7528\u4e34\u5e8a\/\u4eba\u53e3\u5b66\u53d8\u91cf\u548c\u7ec4\u5b66\u6570\u636e\u8fdb\u884c\u751f\u5b58\u5206\u6790\uff0c\u5728\u4ee5\u4e0b\u4ee3\u7801\u4e2d\uff0c\u6211\u4eec\u63d0\u53d6\u4e86\u5973\u6027\u4e73\u817a\u764c\u60a3\u8005\u53ca\u5176\u76f8\u5e94\u7684\u751f\u5b58\u7ed3\u679c\u3001\u4e34\u5e8a\/\u4eba\u53e3\u5b66\u53d8\u91cf\u548cRNA-seq\u7279\u5f81\u3002<\/p>\n\n\n\n

    meta$time <- apply(meta[, c(“days_to_death”, “days_to_last_follow_up”)], 1, max, na.rm = TRUE) \/ 365.25<\/p>\n\n\n\n

    meta$status <- meta$vital_status<\/p>\n\n\n\n

    meta$age <- meta$age_at_diagnosis \/ 365.25<\/p>\n\n\n\n

    clin <- subset(meta, gender == “female” & !duplicated(submitter_id) & time > 0 & !is.na(age))<\/p>\n\n\n\n

    clin <- clin[order(clin$submitter_id), ]<\/p>\n\n\n\n

    RNA_count <- RNA_count[, rownames(clin)]<\/p>\n\n\n\n

    2.\u751f\u5b58\u5206\u6790<\/strong><\/strong><\/p>\n\n\n\n

    \u5bf9\u4e8e\u6211\u4eec\u5728\u4e0a\u4e00\u8282\u4e2d\u63d0\u53d6\u7684TCGA\u4e73\u817a\u764c\u60a3\u8005\u7684\u6570\u636e\uff0c\u53ef\u4ee5\u901a\u8fc7survival\u5305\u4e2d\u7684\u51fd\u6570\u83b7\u5f97\u751f\u5b58\u6982\u7387\u7684Kaplan-Meier\u4f30\u8ba1\u3002<\/p>\n\n\n\n

    # KM\u8bc4\u4f30<\/p>\n\n\n\n

    clin$status[clin$status == “Dead”] <- 1<\/p>\n\n\n\n

    clin$status[clin$status == “Alive”] <- 0<\/p>\n\n\n\n

    clin$status <- as.numeric(clin$status)<\/p>\n\n\n\n

    sfit <- survival::survfit(Surv(time, status) ~ 1, data = clin)<\/p>\n\n\n\n

    # \u8ba1\u7b971\u30013\u30015\u5e74\u65f6\u95f4\u70b9\u7684\u751f\u5b58\u7387<\/p>\n\n\n\n

    summary(sfit, times = c(1, 3, 5))<\/p>\n\n\n\n

    theme_set(theme_bw())<\/p>\n\n\n\n

    ggsurv <- survminer::ggsurvplot(sfit,<\/p>\n\n\n\n

      conf.int = TRUE, risk.table = TRUE,<\/p>\n\n\n\n

      xlab = “Time since diagnosis (year)”,<\/p>\n\n\n\n

      legend = “none”, surv.median.line = “hv”<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    ggsurv$plot <- ggsurv$plot + annotate(“text”, x = 20, y = 0.9, label = “+  Censor”)<\/p>\n\n\n\n

    ggsurv<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    \u6765\u81ea TCGA \u7684 1061 \u4f8b BRCA \u60a3\u8005\u6570\u636e\u7684 Kaplan-Meier \u66f2\u7ebf\u3002\u4ece\u751f\u5b58\u66f2\u7ebf\u4e2d\u53ef\u4ee5\u63ed\u793a\u4e00\u4e9b\u57fa\u672c\u7684\u7edf\u8ba1\u6570\u636e\u3002\u4f8b\u5982\uff0c\u6240\u6709\u4e73\u817a\u764c\u60a3\u8005\u7684\u4f30\u8ba1\u4e2d\u4f4d\u751f\u5b58\u65f6\u95f4\uff0c\u5373\u751f\u5b58\u6982\u7387\u4e3a50%\u7684\u65f6\u95f4\u4e3a10.8\u5e74\uff08\u56fe\u4e2d\u7684\u865a\u7ebf\uff09\uff0c1\u5e74\u30015\u5e74\u548c10\u5e74\u751f\u5b58\u6982\u7387\u5206\u522b\u4e3a0.988\u30010.853\u548c0.658\u3002<\/p>\n\n\n\n

    \u5bf9\u6570\u79e9\u68c0\u9a8c\u53ef\u7528\u4e8e\u6d4b\u8bd5\u4e24\u7ec4\u60a3\u8005 [\u4f8b\u5982\u63a5\u53d7\u6cbb\u7597\uff08\u836f\u7269\/\u653e\u5c04\u6cbb\u7597\uff09\u6216\u672a\u6cbb\u7597] \u662f\u5426\u5177\u6709\u76f8\u540c\uff08\u539f\u5047\u8bbe\uff09\u6216\u4e0d\u540c\u7684\u751f\u5b58\u51fd\u6570\uff08\u66ff\u4ee3\u5047\u8bbe\uff09\uff0c\u5e76\u63d0\u4f9b\u76f8\u5e94\u7684 P \u503c\u3002log-rank \u68c0\u9a8c\u8fd8\u53ef\u7528\u4e8e\u6839\u636e\u5176\u4ed6\u5206\u7c7b\u53d8\u91cf\u6bd4\u8f83\u4efb\u4f55\u60a3\u8005\u4e9a\u7ec4\u7684\u751f\u5b58\u6982\u7387\u3002<\/p>\n\n\n\n

    \u4e3a\u4e86\u6bd4\u8f83\u4e24\u7ec4\u60a3\u8005\u7684\u751f\u5b58\u66f2\u7ebf\uff0c\u4f8b\u5982\u6cbb\u7597\uff08\u5373\u836f\u7269\u6216\u653e\u5c04\u6cbb\u7597\uff09\u6216\u975e\u6cbb\u7597\uff0c\u8be5\u51fd\u6570\u53ef\u4ee5\u6267\u884c\u5bf9\u6570\u79e9\u68c0\u9a8c\u6765\u6bd4\u8f83\u4e24\u6761\u751f\u5b58\u66f2\u7ebf\u3002\u6216\u8005\uff0c\u4f7f\u7528\u5305\u542b\u6cbb\u7597\u7ec4\u4f5c\u4e3a\u534f\u53d8\u91cf\u7684\u516c\u5f0f\u7684\u51fd\u6570\u53ef\u4ee5\u8fd4\u56de\u6bcf\u4e2a\u7ec4\u7684 \uff08KM\uff09 \u751f\u5b58\u6982\u7387\u3002\u7136\u540e\uff0c\u5e26\u6709\u5bf9\u8c61\u7684\u51fd\u6570\u5c06\u7ed8\u5236\u4e24\u6761\u751f\u5b58\u66f2\u7ebf\u5e76\u6267\u884c\u5bf9\u6570\u79e9\u6d4b\u8bd5\uff0c\u5982\u4e0b\u56fe\u6240\u793a\u3002<\/p>\n\n\n\n

    survival::survdiff(Surv(time, status) ~ treatments, data = clin)<\/p>\n\n\n\n

    sfit2 <- survfit(Surv(time, status) ~ treatments, data = clin)<\/p>\n\n\n\n

    ggsurv <- ggsurvplot(sfit2,<\/p>\n\n\n\n

      conf.int = TRUE, risk.table = TRUE,<\/p>\n\n\n\n

      xlab = “Time since diagnosis (year)”, legend = c(.6, .9),<\/p>\n\n\n\n

      legend.labs = c(“No”, “Yes”), legend.title = “Treatment”,<\/p>\n\n\n\n

      risk.table.y.text.col = TRUE, risk.table.y.text = FALSE<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    ggsurv$plot <- ggsurv$plot +<\/p>\n\n\n\n

      annotate(“text”, x = 21, y = 1, label = “+  Censor”) +<\/p>\n\n\n\n

      annotate(“text”, x = 22, y = .88, label = paste0(“Log-rank test:\\n”, surv_pvalue(sfit2)$pval.txt))<\/p>\n\n\n\n

    ggsurv<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    \u6765\u81eaTCGA\u7684BRCA\u60a3\u8005\u751f\u5b58\u6570\u636e\u7684Kaplan-Meier\u66f2\u7ebf\u6309\u6cbb\u7597\uff08\u5373\u836f\u7269\u6216\u653e\u5c04\u6cbb\u7597\uff09\u6216\u975e\u6cbb\u7597\u5206\u7ec4\u3002log-rank \u68c0\u9a8c\u662f\u6bd4\u8f83\u5bf9\u5e94\u4e8e\u4e24\u7ec4\u60a3\u8005\u7684\u4e24\u79cd\u751f\u5b58\u5206\u5e03\u3002<\/p>\n\n\n\n

    \u4e3a\u4e86\u5206\u6790\u8fde\u7eed\u53d8\u91cf\uff08\u4f8b\u5982\u5e74\u9f84\uff09\u662f\u5426\u4e0e\u751f\u5b58\u7ed3\u679c\u76f8\u5173\uff0c\u6211\u4eec\u53ef\u4ee5\u4f7f\u7528\u51fd\u6570\u6765\u62df\u5408 Cox \u6a21\u578b\uff0c\u8be5\u51fd\u6570\u7c7b\u4f3c\u4e8e\u62df\u5408\u7ebf\u6027\u6a21\u578b\u7684\u51fd\u6570\u3002<\/p>\n\n\n\n

    fit_cox <- coxph(Surv(time, status) ~ age, data = clin)<\/p>\n\n\n\n

    summary(fit_cox)<\/p>\n\n\n\n

    Cox \u6a21\u578b\u5047\u8bbe\u534f\u53d8\u91cf\u5177\u6709\u6bd4\u4f8b\u98ce\u9669\u548c\u5bf9\u6570\u7ebf\u6027\u3002\u4e3a\u4e86\u68c0\u67e5\u4e34\u5e8a\u6216\u4eba\u53e3\u7edf\u8ba1\u53d8\u91cf\uff08\u4f8b\u5982\u5e74\u9f84\uff09\u7684\u5bf9\u6570\u7ebf\u6027\uff0c\u6211\u4eec\u53ef\u4ee5\u62df\u5408\u5e74\u9f84\u6548\u5e94\u7684\u60e9\u7f5a\u5e73\u6ed1\u6837\u6761\u3002\u4e0b\u9762\u7684\u4ee3\u7801\u663e\u793a\uff0c\u5e73\u6ed1\u6837\u6761\u7684\u975e\u7ebf\u6027\u90e8\u5206\u5177\u6709\u663e\u8457\u7684\u5f71\u54cd\u3002\u56e0\u6b64\uff0c\u5e74\u9f84\u7684\u5bf9\u6570\u7ebf\u6027\u5047\u8bbe\u4e0d\u6ee1\u8db3\u3002<\/p>\n\n\n\n

    fit_cox_spline <- coxph(Surv(time, status) ~ pspline(age), data = clin)<\/p>\n\n\n\n

    fit_cox_spline<\/p>\n\n\n\n

    \u4e3a\u4e86\u68c0\u67e5\u5e74\u9f84\u7684\u6bd4\u4f8b\u98ce\u9669\uff0c\u6211\u4eec\u53ef\u4ee5\u6dfb\u52a0\u4e00\u4e2a\u968f\u65f6\u95f4\u53d8\u5316\u7684\u534f\u53d8\u91cf\u3002<\/p>\n\n\n\n

    survival::cox.zph(fit_cox, transform = “log”)<\/p>\n\n\n\n

    \u5728\u8fd9\u91cc\uff0c\u68c0\u67e5\u5bf9\u6570\u7ebf\u6027\u6216\u6bd4\u4f8b\u98ce\u9669\u7684\u65b9\u6cd5\u53ea\u80fd\u7528\u4e8e\u4f4e\u7ef4\u6570\u636e\u8bbe\u7f6e\u3002\u5f53\u5305\u62ec\u9ad8\u7ef4\u7ec4\u5b66\u6570\u636e\u65f6\uff0c\u76ee\u524d\u6ca1\u6709\u68c0\u67e5\u5bf9\u6570\u7ebf\u6027\u6216\u6bd4\u4f8b\u98ce\u9669\u7684\u6807\u51c6\u65b9\u6cd5\u3002<\/p>\n\n\n\n

    3. \u751f\u5b58\u6a21\u578b\u9a8c\u8bc1<\/strong><\/strong><\/p>\n\n\n\n

    \u9884\u540e\u6a21\u578b\u7684\u7406\u60f3\u8bc4\u4f30\u662f\u57fa\u4e8e\u5b8c\u5168\u72ec\u7acb\u7684\u9a8c\u8bc1\u6570\u636e\uff0c\u56e0\u4e3a\u57fa\u4e8e\u8bad\u7ec3\u6570\u636e\u6784\u5efa\u7684\u9ad8\u7ef4\u751f\u5b58\u6a21\u578b\u53ef\u80fd\u4f1a\u8fc7\u62df\u5408\u3002\u5982\u679c\u6ca1\u6709\u72ec\u7acb\u7684\u9a8c\u8bc1\u6570\u636e\uff0c\u5efa\u8bae\u4f7f\u7528\u57fa\u4e8e\u91cd\u91c7\u6837\u7684\u65b9\u6cd5\u4f30\u8ba1\u6a21\u578b\u9884\u6d4b\u6027\u80fd\u7684\u4e0d\u786e\u5b9a\u6027\u3002\u4f8b\u5982\uff0c\u53ef\u4ee5\u901a\u8fc7\u53cd\u590d\u5c06\u6570\u636e\u96c6\u62c6\u5206\u4e3a\u8bad\u7ec3\/\u9a8c\u8bc1\u96c6\u5e76\u4f7f\u7528\u5404\u79cd\u8bc4\u4f30\u6307\u6807\u8bc4\u4f30\u6a21\u578b\u5728\u4e0d\u540c\u9a8c\u8bc1\u96c6\u4e0a\u7684\u6027\u80fd\u6765\u5b8c\u6210\u6b64\u64cd\u4f5c\u3002<\/p>\n\n\n\n

    \u7ecf\u5178\u6a21\u578b\u8bc4\u4f30<\/p>\n\n\n\n

    \u4e3a\u4e86\u8bc4\u4f30\u7edf\u8ba1\u6a21\u578b\u7684\u6027\u80fd\uff0c\u6211\u4eec\u9996\u5148\u5c06\u6570\u636e\u62c6\u5206\u4e3a\u8bad\u7ec3\u6570\u636e\u96c6\u548c\u9a8c\u8bc1\u6570\u636e\u96c6\u3002\u4f8b\u5982\uff0c\u6211\u4eec\u53ef\u4ee5\u5c06 TCGA \u7684 1047 \u540d BRCA \u60a3\u8005\u968f\u673a\u62c6\u5206\u4e3a80%\u4f5c\u4e3a\u8bad\u7ec3\u96c6\u548c20%\u4f5c\u4e3a\u9a8c\u8bc1\u96c6\u3002<\/p>\n\n\n\n

    set.seed(123)<\/p>\n\n\n\n

    n <- nrow(x)<\/p>\n\n\n\n

    idx <- sample(1:n, n * 0.8, replace = FALSE)<\/p>\n\n\n\n

    x_train <- x[idx, ]<\/p>\n\n\n\n

    y_train <- y[idx, ]<\/p>\n\n\n\n

    x_validate <- x[-idx, ]<\/p>\n\n\n\n

    y_validate <- y[-idx, ]<\/p>\n\n\n\n

    \u8bc1\u660e\u751f\u5b58\u6a21\u578b\u9884\u540e\u80fd\u529b\u7684\u6700\u7b80\u5355\u65b9\u6cd5\u662f\u5bf9\u9884\u540e\u8bc4\u5206\uff08\u5373\u7ebf\u6027\u9884\u6d4b\u56e0\u5b50\uff09\u8fdb\u884c\u4e8c\u5206\u6cd5LP\u5728Cox\u6a21\u578b\u4e2d\u7684\u4e2d\u503c\uff0c\u7136\u540e\u4f7f\u7528\u5bf9\u6570\u79e9\u68c0\u9a8c\u6765\u6bd4\u8f83\u4e24\u7ec4\u60a3\u8005\u7684\u751f\u5b58\u66f2\u7ebf\u3002<\/p>\n\n\n\n

    # \u8bad\u7ec3Lasso Cox\u6a21\u578b\uff0c\u7c7b\u4f3c\u4e8e\u5176\u4ed6Cox\u7c7b\u578b\u6a21\u578b<\/p>\n\n\n\n

    set.seed(123)<\/p>\n\n\n\n

    cvfit <- cv.glmnet(x_train, y_train, family = “cox”, nfolds = 5, penalty.factor = pf)<\/p>\n\n\n\n

    pred_lp <- predict(cvfit, newx = x_validate, s = cvfit$lambda.min, type = “link”)<\/p>\n\n\n\n

    # \u6839\u636e\u9884\u540e\u8bc4\u5206(\u7ebf\u6027\u9884\u6d4b\u56e0\u5b50)\u4e2d\u4f4d\u6570\u8fdb\u884c\u4e8c\u5206\u7c7b\uff0c\u5c06\u9a8c\u8bc1\u60a3\u8005\u5206\u4e3a\u4e24\u7ec4<\/p>\n\n\n\n

    group_dichotomize <- as.numeric(pred_lp > median(pred_lp))<\/p>\n\n\n\n

    # \u6839\u636eKM\u4f30\u8ba1\u7ed8\u5236\u4e24\u6761\u751f\u5b58\u66f2\u7ebf\uff0c\u5e76\u901a\u8fc7log-rank\u68c0\u9a8c\u8fdb\u884c\u6bd4\u8f83<\/p>\n\n\n\n

    dat_tmp <- data.frame(time = y_validate[, 1], status = y_validate[, 2], group = group_dichotomize)<\/p>\n\n\n\n

    sfit <- survfit(Surv(time, status) ~ group, data = dat_tmp)<\/p>\n\n\n\n

    ggsurv <- ggsurvplot(sfit,<\/p>\n\n\n\n

      conf.int = TRUE, risk.table = TRUE,<\/p>\n\n\n\n

      xlab = “Time since diagnosis (year)”, legend = c(.2, .3),<\/p>\n\n\n\n

      legend.labs = c(“Low risk”, “High risk”), legend.title = “Dichotomized groups”,<\/p>\n\n\n\n

      risk.table.y.text.col = TRUE, risk.table.y.text = FALSE<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    ggsurv$plot <- ggsurv$plot +<\/p>\n\n\n\n

      annotate(“text”, x = 2.6, y = .03, label = paste0(“Log-rank test:\\n”, surv_pvalue(sfit)$pval.txt))<\/p>\n\n\n\n

    ggsurv$table <- ggsurv$table + labs(y = “Dichotomized\\n groups”)<\/p>\n\n\n\n

    ggsurv<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    \u9884\u540e\u8bc4\u5206\u4e5f\u53ef\u4ee5\u6839\u636e\u5206\u4f4d\u6570\u5206\u4e3a\u4e09\u7ec4\u6216\u66f4\u591a\u7ec4\uff0c\u5e76\u4f7f\u7528log-rank\u68c0\u9a8c\u6765\u6bd4\u8f83\u591a\u6761\u751f\u5b58\u66f2\u7ebf\u7684\u5dee\u5f02\u3002<\/p>\n\n\n\n

    group <- pred_lp<\/p>\n\n\n\n

    group[pred_lp >= quantile(pred_lp, 2 \/ 3)] <- 3<\/p>\n\n\n\n

    group[pred_lp >= quantile(pred_lp, 1 \/ 3) & pred_lp < quantile(pred_lp, 2 \/ 3)] <- 2<\/p>\n\n\n\n

    group[pred_lp < quantile(pred_lp, 1 \/ 3)] <- 1<\/p>\n\n\n\n

    #\u6839\u636eKM\u4f30\u8ba1\u7ed8\u5236\u4e24\u6761\u751f\u5b58\u66f2\u7ebf\uff0c\u5e76\u901a\u8fc7log-rank\u68c0\u9a8c\u8fdb\u884c\u6bd4\u8f83<\/p>\n\n\n\n

    dat_tmp <- data.frame(time = y_validate[, 1], status = y_validate[, 2], group = group)<\/p>\n\n\n\n

    sfit <- survfit(Surv(time, status) ~ group, data = dat_tmp)<\/p>\n\n\n\n

    ggsurv <- ggsurvplot(sfit,<\/p>\n\n\n\n

      conf.int = TRUE, risk.table = TRUE,<\/p>\n\n\n\n

      xlab = “Time since diagnosis (year)”, legend = c(.2, .3),<\/p>\n\n\n\n

      legend.labs = c(“Low risk”, “Middle risk”, “High risk”), legend.title = “Groups”,<\/p>\n\n\n\n

      risk.table.y.text.col = TRUE, risk.table.y.text = FALSE<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    ggsurv$plot <- ggsurv$plot +<\/p>\n\n\n\n

      annotate(“text”, x = 3.5, y = .05, label = paste0(“Log-rank test:\\n”, surv_pvalue(sfit)$pval.txt))<\/p>\n\n\n\n

    ggsurv<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    #ROC\u66f2\u7ebf<\/p>\n\n\n\n

    ROC <- risksetROC(<\/p>\n\n\n\n

      Stime = y_validate[, 1], status = y_validate[, 2],<\/p>\n\n\n\n

      marker = pred_lp, predict.time = 5, method = “Cox”,<\/p>\n\n\n\n

      main = “ROC Curve”, col = “seagreen3”, type = “s”,<\/p>\n\n\n\n

      lwd = 2, xlab = “1 – Specificity”, ylab = “Sensitivity”<\/p>\n\n\n\n

    )<\/p>\n\n\n\n

    text(0.7, 0.2, paste(“AUC =”, round(ROC$AUC, 3)))<\/p>\n\n\n\n

    \"\"<\/figure>\n\n\n\n

    \u5c0f\u7ed3<\/strong><\/p>\n\n\n\n

    \u5927\u6d77\u54e5\u5e26\u5927\u5bb6\u5b66\u4e60\u4e86\u4e00\u4e2a\u751f\u5b58\u5206\u6790\u7684\u4e00\u822c\u5de5\u4f5c\u6d41\u7a0b\uff0c\u8be5\u5de5\u4f5c\u6d41\u7a0b\u9002\u7528\u4e8e\u9ad8\u7ef4\u7ec4\u5b66\u6570\u636e\uff0c\u4f5c\u4e3a\u8bc6\u522b\u751f\u5b58\u76f8\u5173\u7279\u5f81\u548c\u9a8c\u8bc1\u751f\u5b58\u6a21\u578b\u65f6\u7684\u8f93\u5165\u3002\u7279\u522b\u662f\uff0c\u6211\u4eec\u91cd\u70b9\u5173\u6ce8\u4e86\u751f\u5b58\u5206\u6790\u4e2d\u5e38\u7528\u7684Cox\u578b\u60e9\u7f5a\u56de\u5f52\u548c\u5206\u5c42\u8d1d\u53f6\u65af\u6a21\u578b\uff0c\u8fd9\u4e9b\u6a21\u578b\u5bf9\u4e8e\u9ad8\u7ef4\u6570\u636e\u7279\u522b\u6709\u7528\u3002\u6700\u540e\u5927\u6d77\u54e5\u7ed9\u5927\u5bb6\u4ecb\u7ecd\u4e00\u4e2a\u4e91\u5de5\u5177\uff01\u540c\u5b66\u4eec\u5982\u679c\u89c9\u5f97\u81ea\u5df1\u7684\u4ee3\u7801\u6c34\u5e73\u4e00\u822c\uff0c\u5bf9\u4e8e\u5f88\u591a\u7684\u53c2\u6570\u4e0d\u77e5\u9053\u600e\u4e48\u6539\uff0c\u53ef\u4ee5\u4f53\u9a8c\u4e00\u4e0b\u6211\u4eec\u7684\u4e91\u751f\u4fe1\u5c0f\u5de5\u5177\uff0c\u53ea\u9700\u8f93\u5165\u6570\u636e\uff0c\u5373\u53ef\u8f7b\u677e\u751f\u6210\u6240\u9700\u56fe\u8868\uff0c\u5b57\u4f53\u5927\u5c0f\u3001\u6807\u9898\u7b49\u4e5f\u53ef\u4e00\u952e\u66f4\u6539\u3002\u611f\u5174\u8da3\u7684\u5c0f\u4f19\u4f34\u53bb\u4e91\u751f\u4fe1\uff08http:\/\/www.biocloudservice.com\/home.html\uff09\u4f53\u9a8c\u4e00\u4e0b\u5427\uff01<\/p>\n","protected":false},"excerpt":{"rendered":"

    \u524d\u8a00 \u751f\u5b58\u5206\u6790\u6587\u7ae0\u4e2d\u6bd4\u6bd4\u7686\u662f\uff0c\u4f60\u7684\u5efa\u6a21\u65b9\u6cd5\u7528\u5bf9\u4e86\u5417\uff1f\u4f60\u7684\u6a21\u578b\u771f\u7684\u51c6\u786e\u5417\uff1f\u662f\u4e0d\u662f\u4e1c\u62fc\u897f\u51d1\u627e\u6765\u4ee3\u7801\uff0c\u53ef\u89c6\u5316\u4e4b\u540e\u8349\u8349 […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"jetpack_post_was_ever_published":false,"_jetpack_newsletter_access":""},"categories":[1],"tags":[],"jetpack_featured_media_url":"","_links":{"self":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/61149"}],"collection":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=61149"}],"version-history":[{"count":1,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/61149\/revisions"}],"predecessor-version":[{"id":61157,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/61149\/revisions\/61157"}],"wp:attachment":[{"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=61149"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=61149"},{"taxonomy":"post_tag","embeddable":true,"href":"http:\/\/www.biocloudservice.com\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=61149"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}