{"id":536,"date":"2023-12-12T20:37:39","date_gmt":"2023-12-12T12:37:39","guid":{"rendered":"https:\/\/blog.ineuro.net\/?p=536"},"modified":"2023-12-12T23:19:17","modified_gmt":"2023-12-12T15:19:17","slug":"%e5%8d%95%e7%bb%86%e8%83%9e%e5%ad%a6%e4%b9%a0%ef%bc%88%e4%ba%94%ef%bc%89%ef%bc%9a%e9%80%9a%e8%bf%87%e8%81%9a%e7%b1%bb%e8%a7%92%e5%ba%a6%e5%88%a4%e6%96%ad%e5%93%aa%e4%ba%9bcluster%e5%8f%af%e8%83%bd","status":"publish","type":"post","link":"https:\/\/blog.ineuro.net\/index.php\/2023\/12\/12\/536\/","title":{"rendered":"\u5355\u7ec6\u80de\u5b66\u4e60\uff08\u4e94\uff09\uff1a\u901a\u8fc7\u805a\u7c7b\u89d2\u5ea6\u5224\u65ad\u54ea\u4e9bcluster\u53ef\u80fd\u5c5e\u4e8e\u540c\u4e00\u7ec6\u80de\u7c7b\u578b"},"content":{"rendered":"\n

1.\u603b\u4f53\u4ee3\u7801<\/h3>\n\n\n\n
##\u6e05\u7a7a\u73af\u5883\nrm(list=ls())\n#\u8bbe\u7f6e\u5de5\u4f5c\u8def\u5f84\n###\u52a0\u8f7d\u6240\u9700\u8981\u7684\u5305\nlibrary(Seurat)\nlibrary(tidyverse)\nlibrary(dplyr)\nlibrary(patchwork)\n\nx=list.files()\n\ndir = c('BC2\/', \"BC21\/\")\nnames(dir) = c('BC2',  'BC21')      \n?Read10X\n\ncounts <- Read10X(data.dir =dir)\nscRNA1 = CreateSeuratObject(counts,min.cells = 3, min.features = 200)\ntable(scRNA1@meta.data$orig.ident)\ndir[1]\n\nscRNAlist <- list()\nfor(i in 1:length(dir)){\n  counts <- Read10X(data.dir = dir[i])\n  scRNAlist[[i]] <- CreateSeuratObject(counts, min.cells = 3, min.features =300)\n}\n\ncounts <- Read10X(data.dir = \"BC2\/\")\nscRNAlist[[1]] <- CreateSeuratObject(counts, min.cells = 3, min.features =300)\n\ncounts <- Read10X(data.dir = \"BC3\/\")\nscRNAlist[[2]] <- CreateSeuratObject(counts, min.cells = 3, min.features =300)\n\ncounts <- Read10X(data.dir = \"BC5\/\")\nscRNAlist[[3]] <- CreateSeuratObject(counts, min.cells = 3, min.features =300)\n\n\n\nsave(scRNAlist,file = \"scRNAlist.Rdata\")\nload(\"F:\/gl\/huada\/scRNAlist.Rdata\")\n\nfor (i in 1:length(scRNAlist)) {\n  scRNAlist[[i]] <- NormalizeData(scRNAlist[[i]])\n  scRNAlist[[i]] <- FindVariableFeatures(scRNAlist[[i]], selection.method = \"vst\",nfeatures = 3000)\n}\n\n\nseur.obj <- scRNAlist[[1]]\n\nseur.obj <- seur.obj %>%\n  Seurat::NormalizeData(verbose = F) %>%\n  FindVariableFeatures(selection.method = 'vst') %>%\n  ScaleData(verbose = F) %>%\n  RunPCA(pc.genes = seur.obj@var.genes,npcs = 100,verbose= F) %>%\n  FindNeighbors(dims= 1:10) %>%\n  FindClusters(resolution= 0.5) %>%\n  RunUMAP(dims= 1:10)\n\nseur.obj2 <- scRNAlist[[2]]\nseur.obj2 <- seur.obj2 %>%\n  Seurat::NormalizeData(verbose = F) %>%\n  FindVariableFeatures(selection.method = 'vst') %>%\n  ScaleData(verbose = F) %>%\n  RunPCA(pc.genes = seur.obj2@var.genes,npcs = 100,verbose= F) %>%\n  FindNeighbors(dims= 1:10) %>%\n  FindClusters(resolution= 0.5) %>%\n  RunUMAP(dims= 1:10)\n\n#\u786e\u4fdd\u4e24\u4e2a\u6570\u636e\u96c6\u7684\u7279\u5f81\u6570\u76f8\u7b49\nshared_genes <- intersect(rownames(seur.obj),row.names(seur.obj2))\n?intersect\n\n#\u8ba1\u7b97cluster\u4e4b\u548c\u5e76\u5f52\u4e00\u5316\n\nseur.obj.matrix <- AggregateExpression(seur.obj,assays = 'RNA',features = shared_genes,slot = 'count')$RNA\n#\u9664\u6bcf\u4e2acluster\u7684\u91cd\u590d\nseur.obj.matrix <- seur.obj.matrix \/ rep(colSums(seur.obj.matrix),each = nrow(seur.obj.matrix))\nseur.obj.matrix <- log10(seur.obj.matrix *100000 +1)\n\nseur.obj.matrix2 <- AggregateExpression(seur.obj2,assays = 'RNA',features = shared_genes,slot = 'count')$RNA\nseur.obj.matrix2 <- seur.obj.matrix2 \/ rep(colSums(seur.obj.matrix2),each = nrow(seur.obj.matrix2))\nseur.obj.matrix2 <- log10(seur.obj.matrix2 *100000 +1)\n\n?AggregateExpression\n?rep\n##\u4e3a\u4e86\u63d0\u9ad8\u51c6\u786e\u6027\uff0c\u5148\u6839\u636e\u76ee\u6807\u6570\u636e\u96c6\u4e2d\u7684\u7ed9\u5b9a\u7ec6\u80de\u7c7b\u578b\u548c\u6574\u4f53\u7ec6\u80de\u7c7b\u578b\u4e2d\u4f4d\u6570\u7684\u500d\u6570\u53d8\u5316\uff0c\u9009\u62e9\u524d200\u4e2a\uff1b\u7136\u540e\u6839\u636e\u76ee\u6807\u6570\u636e\u96c6\u4e2d\u7ed9\u5b9a\u7684\u7ec6\u80de\u7c7b\u578b\u548c\u5176\u4ed6\u7ec6\u80de\u7c7b\u578b\u6700\u5927\u503c\u7684\u500d\u6570\u53d8\u5316\uff0c\u9009\u62e9\u524d200\u4e2a\uff0c\u5bf9\u4e24\u4e2a\u7ed3\u679c\u5408\u5e76\u3002\ncluster <- 3\nseur.obj.gene <- seur.obj.matrix[,cluster + 1]\ngene_fc <- seur.obj.gene \/ apply(seur.obj.matrix,1,median)\ngene_list1 <- names(sort(gene_fc,decreasing = T)[1:200])\n\ngene_fc <- seur.obj.gene \/ apply(seur.obj.matrix[,-(cluster +1)],1,max)\ngene_list2 <- names(sort(gene_fc,decreasing = T)[1:200])\ngene_list <- unique(c(gene_list1,gene_list2))\n\n#3\uff0c\u5728\u76f8\u5173\u7cfb\u6570\u4e0d\u4e3a\u8d1f\u7684\u9650\u5236\u4e0b\uff0c\u6c42Ta\n\nTa <- seur.obj.matrix[gene_list,cluster + 1]\nMb <- seur.obj.matrix2[gene_list,]\ninstall.packages('lsei')\nlibrary(lsei)\nsolv <- nnls(Mb,Ta)\ncorr <- solv$x\ncorr\n\n#\u8c03\u6362\u9884\u6d4b\u6570\u636e\u96c6\u548c\u76ee\u6807\u6570\u636e\u96c6\uff0c\u91cd\u65b0\u8fdb\u884c\u4e0a\u8ff0\u8ba1\u7b97\n# \u6279\u91cf\u8ba1\u7b97\u5404\u4e2a\u7c7b\u7fa4\u7684\u56de\u5f52\u7cfb\u6570\ncluster <- 0\nncol(seur.obj.matrix)\nlist1 <- list()\nfor (cluster in seq(1,ncol(seur.obj.matrix))){\n  seur.obj.gene <- seur.obj.matrix[,cluster]\n  gene_fc <- seur.obj.gene \/ apply(seur.obj.matrix,1,median)\n  gene_list1 <- names(sort(gene_fc,decreasing = T)[1:200])\n  \n  gene_fc <- seur.obj.gene \/ apply(seur.obj.matrix[,-(cluster)],1,max)\n  gene_list2 <- names(sort(gene_fc,decreasing = T)[1:200])\n  gene_list <- unique(c(gene_list1,gene_list2))\n  Ta <- seur.obj.matrix[gene_list,cluster]\n  Mb <- seur.obj.matrix2[gene_list,]\n  solv <- nnls(Mb,Ta)\n  corr <- solv$x\n  list1[[cluster]] <- corr\n}\ncluster <- c()\nlist2 <- list()\nfor (cluster in seq(1,ncol(seur.obj.matrix2))){\n  seur.obj.gene <- seur.obj.matrix2[,cluster]\n  gene_fc <- seur.obj.gene \/ apply(seur.obj.matrix2,1,median)\n  gene_list1 <- names(sort(gene_fc,decreasing = T)[1:200])\n  \n  gene_fc <- seur.obj.gene \/ apply(seur.obj.matrix2[,-(cluster)],1,max)\n  gene_list2 <- names(sort(gene_fc,decreasing = T)[1:200])\n  gene_list <- unique(c(gene_list1,gene_list2))\n  Ta <- seur.obj.matrix2[gene_list,cluster]\n  Mb <- seur.obj.matrix[gene_list,]\n  solv <- nnls(Mb,Ta)\n  corr <- solv$x\n  list2[[cluster]] <- corr\n}\nmat1 <- do.call(rbind,list1)\nmat2 <- do.call(rbind,list2)\n?do.call\nb <- 2*(mat1 + 0.01) * t(mat2 + 0.01)\n?t\nrow.names(b) <- paste0('c',1:nrow(b)-1)\ncolnames(b) <- paste0('c',1:ncol(b)-1)\n\npheatmap::pheatmap(b,cluster_rows = F,cluster_cols = FALSE)\n\nlibrary(pheatmap)\ninstall.packages('psych')\nlibrary(psych)\ncolnames(seur.obj@meta.data)\nIdents(seur.obj) ='Seurat_clusters'\nexp=AverageExpression(seur.obj)\nexp\n?Idents\ncoorda <- corr.test(exp$RNA,exp$RNA,method='spearman')\npheatmap(coorda$r)\n<\/code><\/pre>\n\n\n\n
\u5bf9\u4e8e\u4e0d\u540c\u53d1\u80b2\u65f6\u95f4\uff0c\u5982\u679c\u5f3a\u884c\u6574\u5408\u53ef\u80fd\u5bfc\u81f4\u660e\u663e\u7684\u6279\u6b21\u6548\u5e94\uff0c\u63a9\u76d6\u6700\u521d\u7684\u7ed3\u679c\u3002<\/p>\n\n\n\n
2.\u4ee3\u7801\u8bf4\u660e<\/h3>\n\n\n\n
\uff081\uff09for\u5faa\u73af\u6784\u5efascRNA\u5217\u8868<\/h4>\n\n\n\n
scRNAlist <- list()
for(i in 1:length(dir)){
 counts <- Read10X(data.dir = dir[i])
 scRNAlist[[i]] <- CreateSeuratObject(counts, min.cells = 3, min.features =300)
}<\/p>\n\n\n\n
\u5de5\u4f5c\u73af\u5883\u6587\u6587\u4ef6\u5939\u4e2d\u6709i\u7ec4\u6570\u636e\uff0c\u5c06\u6240\u6709\u6570\u636e\u5408\u5e76\u81f3\u4e00\u4e2alist\uff0c\u5373scRNAlist\u3002<\/p>\n\n\n\n
(2) intersect()<\/h4>\n\n\n\n
\u5728\u4e24\u4e2a\u5411\u91cf\u4e0a\u6267\u884c\u96c6\u5e76\u3001\u4ea4\u3001(\u975e\u5bf9\u79f0!)\u5dee\u3001\u76f8\u7b49\u548c\u96b6\u5c5e\u3002union(x, y)
intersect(x, y)
setdiff(x, y)
setequal(x, y)
is.element(el, set)
#is.element(x, y) is identical to x %in% y.<\/p>\n\n\n\n
(3) AggregateExpression<\/h4>\n\n\n\n
\u8fd4\u56de\u6bcf\u4e2a\u6807\u8bc6\u7c7b\u7684\u805a\u5408(\u6c42\u548c)\u8868\u8fbe\u5f0f\u503cAggregateExpression(
object,
assays = NULL,
features = NULL,
return.seurat = FALSE,
group.by = “ident”,
add.ident = NULL,
slot = “data”,
verbose = TRUE,
…
)
\u200b<\/p>\n\n\n\n
object<\/code><\/th>Seurat object<\/th><\/tr><\/thead>
assays<\/code><\/td>Which assays to use. Default is all assays<\/td><\/tr>
features<\/code><\/td>Features to analyze. Default is all features in the assay<\/td><\/tr>
return.seurat<\/code><\/td>Whether to return the data as a Seurat object. Default is FALSE<\/td><\/tr>
group.by<\/code><\/td>Categories for grouping (e.g, ident, replicate, celltype); ‘ident’ by default<\/td><\/tr>
add.ident<\/code><\/td>(Deprecated) Place an additional label on each cell prior to pseudobulking (very useful if you want to observe cluster pseudobulk values, separated by replicate, for example)<\/td><\/tr>
slot<\/code><\/td>Slot(s) to use; if multiple slots are given, assumed to follow the order of ‘assays’ (if specified) or object’s assays<\/td><\/tr>
verbose<\/code><\/td>Print messages and show progress bar<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
(4) rep()<\/h4>\n\n\n\n
rep\u590d\u5236x\u4e2d\u7684\u503c\u3002\u5b83\u662f\u4e00\u4e2a\u6cdb\u578b\u51fd\u6570\u3002<\/p>\n\n\n\n
each<\/code><\/p>\n\n\n\n
\u975e\u8d1f\u6574\u6570\u3002x\u7684\u6bcf\u4e2a\u5143\u7d20\u6bcf\u6b21\u90fd\u91cd\u590d\u3002\u5176\u4ed6\u8f93\u5165\u5c06\u88ab\u5f3a\u5236\u8f6c\u6362\u4e3a\u6574\u578b\u6216\u53cc\u7cbe\u5ea6\u5411\u91cf\uff0c\u5e76\u83b7\u53d6\u7b2c\u4e00\u4e2a\u5143\u7d20\u3002\u5982\u679cNA\u6216\u65e0\u6548\uff0c\u5219\u89c6\u4e3a1\u3002<\/p>\n\n\n\n
\uff085\uff09nnls()<\/h4>\n\n\n\n
\u5f53\u90e8\u5206\u6216\u5168\u90e8\u89e3\u503c\u53d7\u5230\u975e\u8d1f\u6027\u7ea6\u675f\u65f6\uff0c\u8fd9\u4e9b\u51fd\u6570\u5bf9\u4e8e\u89e3\u51b3\u6700\u5c0f\u4e8c\u4e58\u6216\u4e8c\u6b21\u89c4\u5212\u95ee\u9898\u7279\u522b\u6709\u7528\u3002\u53ef\u4ee5\u8fdb\u4e00\u6b65\u9650\u5236\u795e\u7ecf\u7f51\u7edc\u9650\u5236\u7cfb\u6570\u6709\u4e00\u4e2a\u56fa\u5b9a\u7684\u6b63\u548c\u3002<\/p>\n\n\n\n
(6)do.call()<\/p>\n\n\n\n
\u6839\u636e\u540d\u79f0\u6216\u51fd\u6570\u4ee5\u53ca\u8981\u4f20\u9012\u7ed9\u5b83\u7684\u53c2\u6570\u5217\u8868\u6784\u9020\u5e76\u6267\u884c\u51fd\u6570\u8c03\u7528\u3002do.call(what, args, quote = FALSE, envir = parent.frame())<\/p>\n\n\n\n
\"\u56fe\u7247[1]-\u5403\u4e86\u5403\u4e86\"<\/figure>\n\n\n\n
\"\u56fe\u7247[2]-\u5403\u4e86\u5403\u4e86\"<\/figure>\n\n\n\n
\"\u56fe\u7247[3]-\u5403\u4e86\u5403\u4e86\"<\/figure>\n","protected":false},"excerpt":{"rendered":"
1.\u603b\u4f53\u4ee3\u7801 \u5bf9\u4e8e\u4e0d\u540c\u53d1\u80b2\u65f6\u95f4\uff0c\u5982\u679c\u5f3a\u884c\u6574\u5408\u53ef\u80fd\u5bfc\u81f4\u660e\u663e\u7684\u6279\u6b21\u6548\u5e94\uff0c\u63a9\u76d6\u6700\u521d\u7684\u7ed3\u679c\u3002 2.\u4ee3\u7801\u8bf4\u660e \uff081\uff09fo […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1,24],"tags":[25],"topics":[],"_links":{"self":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/536"}],"collection":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/comments?post=536"}],"version-history":[{"count":3,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/536\/revisions"}],"predecessor-version":[{"id":540,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/536\/revisions\/540"}],"wp:attachment":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/media?parent=536"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/categories?post=536"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/tags?post=536"},{"taxonomy":"topics","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/topics?post=536"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}