{"id":180845,"date":"2024-05-08T20:43:30","date_gmt":"2024-05-08T12:43:30","guid":{"rendered":""},"modified":"2024-05-08T20:43:39","modified_gmt":"2024-05-08T12:43:39","slug":"r%e8%af%ad%e8%a8%80%e6%80%8e%e4%b9%88%e8%ae%be%e7%bd%aefdr%e5%92%8clogfc%e7%9a%84%e5%80%bc%e5%95%8a","status":"publish","type":"post","link":"https:\/\/docs.pingcode.com\/ask\/180845.html","title":{"rendered":"R\u8bed\u8a00\u600e\u4e48\u8bbe\u7f6eFDR\u548clogFC\u7684\u503c\u554a"},"content":{"rendered":"<p style=\"text-align:center\"><img decoding=\"async\" src=\"https:\/\/cdn-kb.worktile.com\/kb\/wp-content\/uploads\/2024\/04\/27070300\/f21a8360-3915-4d83-9dd8-61430cf278f9.webp\" alt=\"R\u8bed\u8a00\u600e\u4e48\u8bbe\u7f6eFDR\u548clogFC\u7684\u503c\u554a\" \/><\/p>\n<p><p><strong>R\u8bed\u8a00\u4e2d\u8bbe\u7f6e\u5047\u53d1\u73b0\u7387\uff08False Discovery Rate, FDR\uff09\u548c\u5bf9\u6570\u500d\u6570\u53d8\u5316\uff08log Fold Change, logFC\uff09\u7684\u503c\uff0c\u901a\u5e38\u6d89\u53ca\u751f\u7269\u7edf\u8ba1\u548c\u57fa\u56e0\u8868\u8fbe\u5206\u6790<\/strong>\u3002\u4f8b\u5982\uff0c\u5728\u4f7f\u7528\u751f\u7269\u4fe1\u606f\u5b66\u5de5\u5177\u5982<code>limma<\/code>\u8fdb\u884c\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0\u5206\u6790\u65f6\uff0c\u9700\u8981\u786e\u5b9a\u54ea\u4e9b\u57fa\u56e0\u5177\u6709\u751f\u7269\u5b66\u4e0a\u663e\u8457\u7684\u8868\u8fbe\u53d8\u5316\u3002\u5728\u8fd9\u4e2a\u573a\u666f\u4e2d\uff0c\u53ef\u4ee5\u901a\u8fc7\u8bbe\u5b9a\u7279\u5b9a\u7684FDR\u9608\u503c\u6765\u63a7\u5236\u5047\u9633\u6027\u7387\uff0c\u540c\u65f6\u5229\u7528logFC\u7684\u503c\u6765\u8bc4\u4f30\u8868\u8fbe\u91cf\u7684\u53d8\u5316\u7a0b\u5ea6\u3002<strong>\u4e00\u822c\u800c\u8a00\uff0c\u901a\u8fc7\u8c03\u6574FDR\u7684\u9608\u503c\u548c\u786e\u5b9alogFC\u7684\u6807\u51c6\uff0c\u7814\u7a76\u4eba\u5458\u80fd\u591f\u7b5b\u9009\u51fa\u5177\u6709\u751f\u7269\u5b66\u610f\u4e49\u7684\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0<\/strong>\u3002<\/p>\n<\/p>\n<p><p>\u63a5\u4e0b\u6765\uff0c\u6211\u4eec\u5c06\u901a\u8fc7\u4e00\u4e2a\u793a\u4f8b\u8be6\u7ec6\u63cf\u8ff0\u5982\u4f55\u5728R\u8bed\u8a00\u4e2d\u8bbe\u7f6eFDR\u548clogFC\u7684\u503c\u3002<\/p>\n<\/p>\n<p><h3>\u4e00\u3001\u5b89\u88c5\u4e0e\u52a0\u8f7d\u5fc5\u8981\u7684\u5305<\/h3>\n<\/p>\n<p><p>\u5728R\u8bed\u8a00\u4e2d\uff0c\u8981\u8fdb\u884c\u5dee\u5f02\u8868\u8fbe\u5206\u6790\uff0c\u6211\u4eec\u9996\u5148\u9700\u8981\u5b89\u88c5\u548c\u52a0\u8f7d\u9002\u7528\u4e8e\u8fd9\u79cd\u5206\u6790\u7684\u5305\uff0c\u5982<code>limma<\/code>\uff1a<\/p>\n<\/p>\n<p><pre><code class=\"language-R\">if (!requireNamespace(&quot;BiocManager&quot;, quietly = TRUE))<\/p>\n<p>    install.packages(&quot;BiocManager&quot;)<\/p>\n<p>BiocManager::install(&quot;limma&quot;)<\/p>\n<p>library(limma)<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><h3>\u4e8c\u3001\u6570\u636e\u51c6\u5907<\/h3>\n<\/p>\n<p><p>\u5dee\u5f02\u8868\u8fbe\u5206\u6790\u7684\u524d\u63d0\u662f\u62e5\u6709\u8868\u8fbe\u91cf\u6570\u636e\uff0c\u901a\u5e38\u662f\u901a\u8fc7RNA\u6d4b\u5e8f\uff08RNA-seq\uff09\u6216\u5fae\u9635\u5217\uff08microarray\uff09\u5b9e\u9a8c\u83b7\u5f97\u3002\u5047\u8bbe\u6211\u4eec\u5df2\u6709\u4e00\u4e2a\u8868\u8fbe\u77e9\u9635<code>exprMatrix<\/code>\u548c\u6837\u672c\u4fe1\u606f<code>sampleInfo<\/code>\uff1a<\/p>\n<\/p>\n<p><pre><code class=\"language-R\"># \u793a\u4f8b\u8868\u8fbe\u77e9\u9635\uff08\u771f\u5b9e\u5e94\u7528\u4e2d\u5e94\u4f7f\u7528\u5b9e\u9a8c\u6570\u636e\uff09<\/p>\n<p>exprMatrix &lt;- matrix(rnorm(1000 * 6), nrow = 1000, ncol = 6)<\/p>\n<p>rownames(exprMatrix) &lt;- paste0(&quot;Gene&quot;, seq(1, 1000))<\/p>\n<p>colnames(exprMatrix) &lt;- paste0(&quot;Sample&quot;, seq(1, 6))<\/p>\n<h2><strong>\u793a\u4f8b\u6837\u672c\u4fe1\u606f<\/strong><\/h2>\n<p>sampleInfo &lt;- data.frame(<\/p>\n<p>    Group = factor(c(&quot;Control&quot;, &quot;Control&quot;, &quot;Control&quot;, &quot;Treatment&quot;, &quot;Treatment&quot;, &quot;Treatment&quot;))<\/p>\n<p>)<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><h3>\u4e09\u3001\u6784\u5efa\u8bbe\u8ba1\u77e9\u9635\u4e0e\u6bd4\u8f83<\/h3>\n<\/p>\n<p><p>\u4f7f\u7528<code>model.matrix<\/code>\u6784\u5efa\u8bbe\u8ba1\u77e9\u9635\uff0c\u7136\u540e\u4f7f\u7528<code>makeContrasts<\/code>\u5b9a\u4e49\u6bd4\u8f83\uff1a<\/p>\n<\/p>\n<p><pre><code class=\"language-R\">design &lt;- model.matrix(~0 + Group, data = sampleInfo)<\/p>\n<p>colnames(design) &lt;- levels(sampleInfo$Group)<\/p>\n<p>contrast.matrix &lt;- makeContrasts(TreatmentVsControl = Treatment - Control, levels = design)<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><h3>\u56db\u3001\u62df\u5408\u6a21\u578b\u4e0e\u83b7\u53d6\u7edf\u8ba1\u91cf<\/h3>\n<\/p>\n<p><p>\u4f7f\u7528<code>lmFit<\/code>\u548c<code>contrasts.fit<\/code>\u62df\u5408\u6a21\u578b\uff0c\u5e76\u901a\u8fc7<code>eBayes<\/code>\u8ba1\u7b97\u7edf\u8ba1\u91cf\uff1a<\/p>\n<\/p>\n<p><pre><code class=\"language-R\">fit &lt;- lmFit(exprMatrix, design)<\/p>\n<p>fit2 &lt;- contrasts.fit(fit, contrast.matrix)<\/p>\n<p>fit2 &lt;- eBayes(fit2)<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><h3>\u4e94\u3001\u8bbe\u7f6eFDR\u548clogFC\u9608\u503c\u8fdb\u884c\u7b5b\u9009<\/h3>\n<\/p>\n<p><p>\u73b0\u5728\uff0c\u6211\u4eec\u53ef\u4ee5\u6839\u636e\u8bbe\u5b9a\u7684FDR\u548clogFC\u9608\u503c\u6765\u7b5b\u9009\u663e\u8457\u7684\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0\uff1a<\/p>\n<\/p>\n<p><pre><code class=\"language-R\"># \u8bbe\u7f6eFDR\u9608\u503c\u4e3a0.05\uff0clogFC\u9608\u503c\u4e3a1<\/p>\n<p>fdrThreshold &lt;- 0.05<\/p>\n<p>logFCThreshold &lt;- 1<\/p>\n<h2><strong>\u4f7f\u7528decideTests\u51fd\u6570\u83b7\u53d6\u7b26\u5408\u6761\u4ef6\u7684\u57fa\u56e0<\/strong><\/h2>\n<p>decisions &lt;- decideTests(fit2, adjust.method = &quot;BH&quot;, p.value = fdrThreshold)<\/p>\n<h2><strong>\u7b5b\u9009\u51fa\u663e\u8457\u7684\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0<\/strong><\/h2>\n<p>significantGenes &lt;- fit2$genes[as.logical(decisions), ]<\/p>\n<h2><strong>\u8fdb\u4e00\u6b65\u7b5b\u9009\u51fa\u6ee1\u8db3logFC\u9608\u503c\u7684\u57fa\u56e0<\/strong><\/h2>\n<p>sigGenesWithLogFC &lt;- significantGenes[abs(significantGenes$logFC) &gt;= logFCThreshold, ]<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><h3>\u516d\u3001\u7ed3\u679c\u5904\u7406\u4e0e\u89e3\u91ca<\/h3>\n<\/p>\n<p><p>\u5728\u5f97\u5230\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0\u5217\u8868\u540e\uff0c\u6211\u4eec\u53ef\u4ee5\u5bf9\u7ed3\u679c\u8fdb\u884c\u8fdb\u4e00\u6b65\u5904\u7406\u548c\u89e3\u91ca\uff0c\u4f8b\u5982\u8fdb\u884c\u6ce8\u91ca\u548c\u529f\u80fd\u5206\u6790\u3002<\/p>\n<\/p>\n<p><pre><code class=\"language-R\"># \u67e5\u770b\u7b26\u5408\u7b5b\u9009\u6807\u51c6\u7684\u57fa\u56e0\u6570\u91cf<\/p>\n<p>nrow(sigGenesWithLogFC)<\/p>\n<h2><strong>\u5bfc\u51fa\u663e\u8457\u7684\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0\u5217\u8868<\/strong><\/h2>\n<p>write.csv(sigGenesWithLogFC, file = &quot;SignificantGenes.csv&quot;)<\/p>\n<p><\/code><\/pre>\n<\/p>\n<p><p>\u5728\u5e94\u7528\u4e0a\u8ff0\u6b65\u9aa4\u65f6\uff0c\u7814\u7a76\u4eba\u5458\u5e94\u6839\u636e\u5177\u4f53\u7684\u5b9e\u9a8c\u8bbe\u8ba1\u548c\u7814\u7a76\u76ee\u6807\u6765\u8c03\u6574FDR\u548clogFC\u7684\u9608\u503c\u3002<strong>\u8fd9\u6837\u7684\u5206\u6790\u80fd\u5e2e\u52a9\u7814\u7a76\u4eba\u5458\u8bc6\u522b\u90a3\u4e9b\u5728\u7edf\u8ba1\u4e0a\u6709\u663e\u8457\u53d8\u5316\u4e14\u751f\u7269\u5b66\u4e0a\u53ef\u80fd\u91cd\u8981\u7684\u57fa\u56e0<\/strong>\uff0c\u4ece\u800c\u4e3a\u6df1\u5316\u751f\u7269\u5b66\u673a\u5236\u7684\u7406\u89e3\u548c\u672a\u6765\u7684\u7814\u7a76\u65b9\u5411\u63d0\u4f9b\u91cd\u8981\u7ebf\u7d22\u3002<\/p>\n<\/p>\n<h2><strong>\u76f8\u5173\u95ee\u7b54FAQs\uff1a<\/strong><\/h2>\n<p><strong>Q1: \u5728R\u8bed\u8a00\u4e2d\uff0c\u5982\u4f55\u8bbe\u7f6eFDR\uff08False Discovery Rate\uff09\u7684\u503c\uff1f<\/strong><\/p>\n<p>\u5728R\u8bed\u8a00\u4e2d\uff0c\u8bbe\u7f6eFDR\u503c\u53ef\u4ee5\u4f7f\u7528<code>p.adjust<\/code>\u51fd\u6570\u3002\u9996\u5148\uff0c\u5c06\u9700\u8981\u8fdb\u884c\u591a\u91cd\u68c0\u9a8c\u7684p-value\u5b58\u50a8\u5728\u4e00\u4e2a\u53d8\u91cf\u4e2d\uff0c\u7136\u540e\u4f7f\u7528<code>p.adjust<\/code>\u51fd\u6570\u5c06\u5176\u8fdb\u884c\u6821\u6b63\u3002\u53ef\u4ee5\u6307\u5b9a\u4e0d\u540c\u7684\u65b9\u6cd5\u6765\u8ba1\u7b97FDR\uff0c\u5982Benjamini-Hochberg\u65b9\u6cd5\uff0cBonferroni\u65b9\u6cd5\u7b49\u3002\u4f8b\u5982\uff0c\u4f7f\u7528Benjamini-Hochberg\u65b9\u6cd5\u53ef\u4ee5\u6309\u7167\u4ee5\u4e0b\u6b65\u9aa4\u8fdb\u884c\u8bbe\u7f6e\uff1a<\/p>\n<pre><code class=\"language-R\"># \u5047\u8bbep-values\u5b58\u50a8\u5728\u53d8\u91cfp\u4e2d\nadjusted_p &lt;- p.adjust(p, method = &quot;BH&quot;)\n<\/code><\/pre>\n<p>\u8fd9\u5c06\u751f\u6210\u6821\u6b63\u540e\u7684FDR\u503c\uff0c\u5b58\u50a8\u5728<code>adjusted_p<\/code>\u53d8\u91cf\u4e2d\u3002<\/p>\n<p><strong>Q2: \u5728R\u8bed\u8a00\u4e2d\uff0c\u5982\u4f55\u8bbe\u7f6elogFC\uff08log Fold Change\uff09\u7684\u503c\uff1f<\/strong><\/p>\n<p>\u5728R\u8bed\u8a00\u4e2d\uff0c\u8bbe\u7f6elogFC\u503c\u901a\u5e38\u662f\u5728\u57fa\u56e0\u8868\u8fbe\u5206\u6790\u4e2d\u5e94\u7528\u3002\u5047\u8bbe\u4f60\u6709\u4e24\u4e2a\u6761\u4ef6\uff08\u5982\u5bf9\u7167\u7ec4\u548c\u5b9e\u9a8c\u7ec4\uff09\u7684\u57fa\u56e0\u8868\u8fbe\u6570\u636e\uff0c\u53ef\u4ee5\u4f7f\u7528\u4ee5\u4e0b\u6b65\u9aa4\u8ba1\u7b97logFC\u503c\uff1a<\/p>\n<ol>\n<li>\u9996\u5148\uff0c\u8ba1\u7b97\u5b9e\u9a8c\u7ec4\u548c\u5bf9\u7167\u7ec4\u7684\u8868\u8fbe\u91cf\u5dee\u5f02\uff0c\u4e5f\u5c31\u662fraw fold change\uff08FC\uff09\u503c\u3002\u8fd9\u53ef\u4ee5\u901a\u8fc7\u5c06\u5b9e\u9a8c\u7ec4\u548c\u5bf9\u7167\u7ec4\u7684\u8868\u8fbe\u91cf\u76f8\u9664\u6765\u5b8c\u6210\u3002<\/li>\n<li>\u4f7f\u7528<code>log2<\/code>\u51fd\u6570\u5c06\u8be5\u503c\u8f6c\u6362\u4e3alog2\u7684\u5f62\u5f0f\u3002<\/li>\n<\/ol>\n<p>\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\uff1a<\/p>\n<pre><code class=\"language-R\"># \u5047\u8bbegene1_control\u548cgene1_experiment\u5206\u522b\u662f\u5bf9\u7167\u7ec4\u548c\u5b9e\u9a8c\u7ec4\u7684\u57fa\u56e0\u8868\u8fbe\u6570\u636e\nFC &lt;- gene1_experiment \/ gene1_control\nlogFC &lt;- log2(FC)\n<\/code><\/pre>\n<p>\u8fd9\u5c06\u7ed9\u51falogFC\u503c\u3002<\/p>\n<p><strong>Q3: R\u8bed\u8a00\u4e2d\u5982\u4f55\u540c\u65f6\u8bbe\u7f6eFDR\u548clogFC\u7684\u503c\uff1f<\/strong><\/p>\n<p>\u5728\u57fa\u56e0\u8868\u8fbe\u5206\u6790\u4e2d\uff0c\u5e38\u5e38\u9700\u8981\u540c\u65f6\u8bbe\u7f6eFDR\u548clogFC\u7684\u9608\u503c\uff0c\u4ee5\u8fdb\u884c\u5dee\u5f02\u8868\u8fbe\u57fa\u56e0\u7684\u7b5b\u9009\u3002\u53ef\u4ee5\u4f7f\u7528\u5982\u4e0b\u65b9\u6cd5\u5b9e\u73b0\uff1a<\/p>\n<ol>\n<li>\u9996\u5148\uff0c\u6839\u636e\u5b9e\u9a8c\u7684\u9700\u6c42\uff0c\u786e\u5b9a\u6240\u9700\u7684FDR\u548clogFC\u7684\u9608\u503c\u3002<\/li>\n<li>\u5c06FDR\u548clogFC\u7684\u9608\u503c\u5206\u522b\u5e94\u7528\u4e8e\u76f8\u5e94\u7684\u53d8\u91cf\u6216\u6570\u636e\u3002<\/li>\n<li>\u4f7f\u7528\u6761\u4ef6\u7b5b\u9009\u6765\u9009\u62e9\u7b26\u5408\u6307\u5b9a\u9608\u503c\u7684\u57fa\u56e0\u3002<\/li>\n<\/ol>\n<p>\u4ee5\u4e0b\u662f\u4e00\u4e2a\u793a\u4f8b\uff1a<\/p>\n<pre><code class=\"language-R\"># \u5047\u8bbep-values\u5b58\u50a8\u5728\u53d8\u91cfp\u4e2d\uff0clogFC\u5b58\u50a8\u5728\u53d8\u91cflogFC\u4e2d\nthreshold_FDR &lt;- 0.05\nthreshold_logFC &lt;- 1\n\n# \u5e94\u7528FDR\u9608\u503c\nadjusted_p &lt;- p.adjust(p, method = &quot;BH&quot;)\nsignificant_genes &lt;- p &lt; threshold_FDR\n\n# \u5e94\u7528logFC\u9608\u503c\nsignificant_genes &lt;- significant_genes &amp; abs(logFC) &gt; threshold_logFC\n<\/code><\/pre>\n<p>\u4ee5\u4e0a\u4ee3\u7801\u5c06\u7b5b\u9009\u51faFDR\u5c0f\u4e8e0.05\u4e14logFC\u7edd\u5bf9\u503c\u5927\u4e8e1\u7684\u663e\u8457\u57fa\u56e0\u3002<\/p>\n","protected":false},"excerpt":{"rendered":"R\u8bed\u8a00\u4e2d\u8bbe\u7f6e\u5047\u53d1\u73b0\u7387\uff08False Discovery Rate, FDR\uff09\u548c\u5bf9\u6570\u500d\u6570\u53d8\u5316\uff08log Fold C [&hellip;]","protected":false},"author":3,"featured_media":180858,"comment_status":"closed","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[37],"tags":[],"acf":[],"_links":{"self":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/180845"}],"collection":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/comments?post=180845"}],"version-history":[{"count":0,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/posts\/180845\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media\/180858"}],"wp:attachment":[{"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/media?parent=180845"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/categories?post=180845"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/docs.pingcode.com\/wp-json\/wp\/v2\/tags?post=180845"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}