{"id":444,"date":"2023-08-28T23:15:17","date_gmt":"2023-08-28T15:15:17","guid":{"rendered":"https:\/\/blog.ineuro.net\/?p=444"},"modified":"2023-09-08T13:26:25","modified_gmt":"2023-09-08T05:26:25","slug":"444","status":"publish","type":"post","link":"https:\/\/blog.ineuro.net\/index.php\/2023\/08\/28\/444\/","title":{"rendered":"\u5355\u7ec6\u80de\u6d4b\u5e8f\uff08\u56db\uff09\uff1a\u8f6f\u4ef6\u6ce8\u91ca"},"content":{"rendered":"

\u4e00\u3001singleR<\/h1>\n

1. \u603b\u4f53\u4ee3\u7801<\/h2>\n
#\u7b2c\u4e8c\u79cd\u65b9\u6cd5\u7528SingleR\u9274\u5b9a\u7ec6\u80de\u7c7b\u578b\n\n###\u4e0b\u8f7d\u597d\u6570\u636e\u5e93\u540e\uff0c\u628aref_Human_all.Rdata\u52a0\u8f7d\u5230\u73af\u5883\u4e2d\uff0c\u8fd9\u6837\u7b97\u662f\u5bf9\u6570\u636e\u5e93\u7684\u52a0\u8f7d\uff0c\u5c31\u53ef\u4ee5\u6309\u7167singler\u7684\u7b97\u6cd5\u6765\u5bf9\u7ec6\u80de\u4e9a\u7fa4\u8fdb\u884c\u5b9a\u4e49\u4e86\u3002\nload(\"~\/ref_Human_all.RData\")\n###\u6211\u4eec\u53ef\u4ee5\u770b\u5230\u5728\u73af\u5883\u4e2d\u591a\u4e86\u4e00\u4e2a\u53ebref_Human_all\u7684\u6587\u4ef6 \u5927\u5c0f\u4e3a113mb  \u8fd9\u4e2a\u5c31\u662f\u6570\u636e\u5e93\n####\u7136\u540e\u6211\u4eec\u628a\u73af\u5883\u4e2d\u7684ref_Human_all\u8d4b\u503c\u4e0erefdata\nrefdata <- ref_Human_all\n###\u628arna\u7684\u8f6c\u5f55\u8868\u8fbe\u6570\u636e\u63d0\u53d6\n?GetAssayData\ntestdata <- GetAssayData(scRNA_harmony, slot=\"data\")\n###\u628ascRNA\u6570\u636e\u4e2d\u7684seurat_clusters\u63d0\u53d6\u51fa\u6765\uff0c\u6ce8\u610f\u8fd9\u91cc\u662f\u56e0\u5b50\u7c7b\u578b\u7684\nclusters <- scRNA_harmony@meta.data$seurat_clusters\n###\u5f00\u59cb\u7528singler\u5206\u6790\ncellpred <- SingleR(test = testdata, ref = refdata, labels = refdata$label.main, \n                    method = \"cluster\", clusters = clusters, \n                    assay.type.test = \"logcounts\", assay.type.ref = \"logcounts\")\n###\u5236\u4f5c\u7ec6\u80de\u7c7b\u578b\u7684\u6ce8\u91ca\u6587\u4ef6\ncelltype = data.frame(ClusterID=rownames(cellpred), celltype=cellpred$labels, stringsAsFactors = FALSE)\n###\u4fdd\u5b58\u4e00\u4e0b\nwrite.csv(celltype,\"celltype_singleR.csv\",row.names = FALSE)\n##\u628asingler\u7684\u6ce8\u91ca\u5199\u5230metadata\u4e2d \u6709\u4e24\u79cd\u65b9\u6cd5\n###\u65b9\u6cd5\u4e00\nscRNA_harmony@meta.data$celltype <- NA\ncelltype$ClusterID <- as.integer(celltype$ClusterID)\nscRNA_harmony@meta.data$seurat_clusters1 <- scRNA_harmony@meta.data$seurat_clusters\nscRNA_harmony@meta.data$seurat_clusters1 <- as.integer(scRNA_harmony@meta.data$seurat_clusters1)\n?which\nfor(i in 1:nrow(celltype)){\n  scRNA_harmony@meta.data[which(scRNA_harmony@meta.data$seurat_clusters == celltype$ClusterID[i]),'celltype'] <- celltype$celltype[i]\n}\n\n###\u56e0\u4e3a\u6211\u628asingler\u7684\u6ce8\u91ca\u52a0\u8f7d\u5230metadata\u4e2d\u65f6\u5019\uff0c\u547d\u540d\u7684\u540d\u5b57\u53ebcelltype\uff0c\u6240\u4ee5\u753b\u56fe\u65f6\u5019\uff0cgroup.by=\"celltype\"\nDimPlot(scRNA_harmony, group.by=\"celltype\", label=T, label.size=5)\n###\u65b9\u6cd5\u4e8c\uff1a\ncelltype = data.frame(ClusterID=rownames(cellpred), celltype=cellpred$labels, stringsAsFactors = F) \nscRNA_harmony@meta.data$singleR=celltype[match(clusters,celltype$ClusterID),'celltype']\n###\u56e0\u4e3a\u6211\u628asingler\u7684\u6ce8\u91ca\u52a0\u8f7d\u5230metadata\u4e2d\u65f6\u5019\uff0c\u547d\u540d\u7684\u540d\u5b57\u53ebsingleR\uff0c\u6240\u4ee5\u753b\u56fe\u65f6\u5019\uff0cgroup.by=\"singleR\"\nDimPlot(scRNA_harmony, group.by=\"singleR\", label=T, label.size=5)\n\n#####\u4f7f\u7528\u6765\u81eascRNAseq\u5305\u4e2d\u7684\u4e24\u4e2a\u4eba\u7c7b\u80f0\u817a\u6570\u636e\u96c6\u3002\u76ee\u7684\u662f\u4f7f\u7528\u4e00\u4e2a\u9884\u5148\u6807\u8bb0\u597d\u7684\u6570\u636e\u96c6\u5bf9\u53e6\u4e00\u4e2a\u672a\u6807\u8bb0\u7684\u6570\u636e\u96c6\u8fdb\u884c\u7ec6\u80de\u7c7b\u578b\u6ce8\u91ca\u3002\n#\u9996\u5148\uff0c\u6211\u4eec\u4f7f\u7528Muraro et al.(2016)\u7684\u6570\u636e\u4f5c\u4e3a\u6211\u4eec\u7684\u53c2\u8003\u6570\u636e\u96c6\u3002\nBiocManager::install(\"scRNAseq\")\nlibrary(scRNAseq)\nsceM <- MuraroPancreasData()\nsceM.1 <- sceM[,!is.na(sceM$label)]\nBiocManager::install(\"scater\")\nlibrary(scater)\nsceM.1 <- logNormCounts(sceM.1)\n##\u63a5\u4e0b\u6765\uff0c\u6211\u4eec\u4f7f\u7528Grun et al.(2016)\u7684\u6570\u636e\u4f5c\u4e3a\u6d4b\u8bd5\u6570\u636e\u96c6\u3002\n##\u6709\u65f6\u5019\u53ef\u80fd\u7f51\u901f\u4e0d\u597d\u54df\nsceG <- GrunPancreasData()\n\n# Remove libraries with no counts.\nsceG <- sceG[,colSums(counts(sceG)) > 0] \nsceG <- logNormCounts(sceG) \n#\u4e3a\u4e86\u52a0\u5feb\u5206\u6790\u7684\u901f\u5ea6\uff0c\u6211\u4eec\u6311\u9009\u524d200\u4e2a\u7ec6\u80de\u8fdb\u884c\u5206\u6790\u3002\nsceG <- sceG[,1:500]\n# \u4f7f\u7528SingleR\u51fd\u6570\u8fdb\u884c\u7ec6\u80de\u7c7b\u578b\u6ce8\u91ca\uff0c\u5e76\u6307\u5b9ade.method=\"wilcox\"\u68c0\u6d4b\u65b9\u6cd5\npred.grun <- SingleR(test=sceG, ref=sceM.1, labels=sceM.1$label, de.method=\"wilcox\")\n\n# \u67e5\u770b\u7ec6\u80de\u7c7b\u578b\u6ce8\u91ca\u7684\u9884\u6d4b\u7ed3\u679c\ntable(pred.grun$labels)\n\n####\u6ce8\u91ca\u7ed3\u679c\u8bca\u65ad\n###1.\u57fa\u4e8escores within cells\n### Annotation diagnostics\nplotScoreHeatmap(pred.grun)\n# \u57fa\u4e8e per-cell \u201cdeltas\u201d\u8bca\u65ad\uff0cDelta\u503c\u4f4e\uff0c\u8bf4\u660e\u6ce8\u91ca\u7ed3\u679c\u4e0d\u662f\u5f88\u660e\u786e\nplotDeltaDistribution(pred.hesc, ncol = 3)\n\n<\/code><\/pre>\n
2. GetAssayData()<\/h2>\n
Assays\u5bf9\u8c61\u7684\u901a\u7528\u8bbf\u95ee\u51fd\u6570\u548c\u8bbe\u7f6e\u51fd\u6570\u3002GetAssayData \u53ef\u7528\u4e8e\u4ece\u4efb\u4f55\u8868\u8fbe\u5f0f\u77e9\u9635\uff08\u5982 "counts"\u3001"data "\u6216 "scale.data"\uff09\u4e2d\u63d0\u53d6\u4fe1\u606f\u3002SetAssayData \u53ef\u7528\u4e8e\u66ff\u6362\u5176\u4e2d\u4e00\u4e2a\u8868\u8fbe\u5f0f\u77e9\u9635<\/p>\n
3. SingleR()<\/h2>\n
\u6839\u636e\u540c\u4e00\u7279\u5f81\u7a7a\u95f4\u4e2d\u5df2\u6807\u6ce8\u7684\u53c2\u8003\u6570\u636e\u96c6\uff0c\u8fd4\u56de\u6d4b\u8bd5\u6570\u636e\u96c6\u4e2d\u6bcf\u4e2a\u5355\u5143\u683c\u7684\u6700\u4f73\u6ce8\u91ca\u3002SingleR\u662f\u4e00\u79cd\u7528\u4e8e\u5355\u7ec6\u80deRNA\u6d4b\u5e8f(scRNAseq)\u6570\u636e\u7684\u81ea\u52a8\u6807\u6ce8\u65b9\u6cd5\u3002\u7ed9\u5b9a\u4e00\u4e2a\u5177\u6709\u5df2\u77e5\u6807\u7b7e\u7684\u53c2\u8003\u6837\u672c\u96c6(\u5355\u7ec6\u80de\u6216\u6279\u91cf)\uff0c\u5b83\u6839\u636e\u4e0e\u53c2\u8003\u7684\u76f8\u4f3c\u6027\u5bf9\u6765\u81ea\u6d4b\u8bd5\u6570\u636e\u96c6\u7684\u65b0\u5355\u5143\u683c\u8fdb\u884c\u6807\u8bb0\u3002\u56e0\u6b64\uff0c\u5bf9\u4e8e\u53c2\u8003\u6570\u636e\u96c6\uff0c\u624b\u52a8\u89e3\u91ca\u96c6\u7fa4\u548c\u5b9a\u4e49\u6807\u8bb0\u57fa\u56e0\u7684\u8d1f\u62c5\u53ea\u9700\u8981\u505a\u4e00\u6b21\uff0c\u5e76\u4e14\u8fd9\u79cd\u751f\u7269\u77e5\u8bc6\u53ef\u4ee5\u4ee5\u81ea\u52a8\u7684\u65b9\u5f0f\u4f20\u64ad\u5230\u65b0\u7684\u6570\u636e\u96c6\u3002singleR\u81ea\u5e26\u76847\u4e2a\u53c2\u8003\u6570\u636e\u96c6\uff0c\u5176\u4e2d5\u4e2a\u662f\u4eba\u7c7b\u6570\u636e\uff0c2\u4e2a\u662f\u5c0f\u9f20\u7684\u6570\u636e\uff1a<\/p>\n
    \n
  1. BlueprintEncodeData Blueprint (Martens and Stunnenberg 2013) and Encode (The ENCODE Project Consortium 2012) \uff08\u4eba\uff09<\/li>\n
  2. DatabaseImmuneCellExpressionData The Database for Immune Cell Expression(\/eQTLs\/Epigenomics)(Schmiedel et al. 2018)\uff08\u4eba\uff09<\/li>\n
  3. HumanPrimaryCellAtlasData the Human Primary Cell Atlas (Mabbott et al. 2013)\uff08\u4eba\uff09<\/li>\n
  4. MonacoImmuneData, Monaco Immune Cell Data - GSE107011 (Monaco et al. 2019)\uff08\u4eba\uff09<\/li>\n
  5. NovershternHematopoieticData Novershtern Hematopoietic Cell Data - GSE24759\uff08\u4eba\uff09<\/li>\n
  6. ImmGenData the murine ImmGen (Heng et al. 2008) \uff08\u9f20\uff09<\/li>\n
  7. MouseRNAseqData a collection of mouse data sets downloaded from GEO (Benayoun et al. 2019).\uff08\u9f20\uff09<\/li>\n<\/ol>\n
    4.plotScoreHeatmap()<\/h2>\n
    plotScoreHeatmap\u663e\u793a\u6240\u6709\u5f15\u7528\u7ec6\u80de\u7c7b\u578b\u4e0a\u6240\u6709\u7ec6\u80de\u7684\u5f97\u5206\uff0c\u8fd9\u5141\u8bb8\u7528\u6237\u68c0\u67e5\u6570\u636e\u96c6\u4e2d\u9884\u6d4b\u7ec6\u80de\u7c7b\u578b\u7684\u53ef\u4fe1\u5ea6\u3002\u6bcf\u4e2a\u7c7b\u7fa4\/\u7ec6\u80de\u7684\u5b9e\u9645\u5206\u914d\u6807\u7b7e\u663e\u793a\u5728\u9876\u90e8\u7684\u989c\u8272\u680f\u4e2d\u3002\u5173\u952e\u70b9\u662f\u68c0\u67e5\u5206\u6570(scores)\u5728\u6bcf\u4e2a\u7c7b\u7fa4\/\u7ec6\u80de\u4e2d\u7684\u5206\u5e03\u60c5\u51b5\u3002\u7406\u60f3\u60c5\u51b5\u4e0b\uff0c\u6bcf\u4e2a\u7c7b\u7fa4\/\u7ec6\u80de(\u5373\u70ed\u56fe\u7684\u5217)\u5e94\u8be5\u6709\u4e00\u4e2a\u660e\u663e\u5927\u4e8e\u5176\u4ed6\u7ec6\u80de\u7684\u5206\u6570\uff0c\u8fd9\u8868\u660e\u5b83\u660e\u786e\u5730\u5206\u914d\u7ed9\u4e86\u5355\u4e2a\u6807\u7b7e\u3002<\/p>\n
    \u4e8c\u3001Garnett<\/h1>\n
    \u4ece\u5355\u7ec6\u80de\u8868\u8fbe\u6570\u636e\u4e2d\u5b9e\u73b0\u81ea\u52a8\u7ec6\u80de\u7c7b\u578b\u5206\u7c7b\u7684\u8f6f\u4ef6\u5305\u3002\u4f7f\u7528\u4eba\u5de5\u5b9a\u4e49\u7684marker\u57fa\u56e0\u9009\u62e9\u7ec6\u80de\uff0c\u57fa\u4e8e\u8fd9\u4e9b\u7ec6\u80de\u4f7f\u7528\u5f39\u6027\u7f51\u7edc\u56de\u5f52\u7684\u673a\u5668\u5b66\u4e60\u7b97\u6cd5\u8bad\u7ec3\u5206\u7c7b\u5668\u3002<\/p>\n
    \u81ea\u5df1\u8bad\u7ec3\u5206\u7c7b\u5668\uff1a<\/p>\n
    1.\u5236\u4f5c\u7b26\u5408garnett\u683c\u5f0f\u8981\u6c42\u7684\u5b9a\u4e49\u7ec6\u80de\u7c7b\u578b\u7684marker\u57fa\u56e0\u6587\u672c\u6587\u4ef6\uff08marker file\uff09\uff1b<\/p>\n
    2.\u4f7f\u7528\u5355\u7ec6\u80de\u6570\u636e\u521b\u5efamonocle3\u7684CDS\u6570\u636e\u5bf9\u8c61\uff1b<\/p>\n
    3.\u5c06\u4e24\u79cd\u6587\u4ef6\u8f93\u5165garnett\uff0c\u5bf9marker\u57fa\u56e0\u8fdb\u884c\u6253\u5206\uff0c\u6839\u636e\u8bc4\u5206\u4f18\u5316marker file\uff1b<\/p>\n
    4.\u5c06\u4f18\u5316\u540e\u7684marker file\u548ccds object\u8f93\u5165garnett\u8bad\u7ec3\u5206\u7c7b\u5668\u3002<\/p>\n
    BiocManager::install(c('BiocGenerics', 'DelayedArray', 'DelayedMatrixStats',\n                       'limma', 'S4Vectors', 'SingleCellExperiment',\"batchelor\",\n                       'SummarizedExperiment'))\n\n# \u5b89\u88c5monocle3\ndevtools::install_github('cole-trapnell-lab\/monocle3')\nlibrary(monocle3)\n## \u7136\u540e\u5b89\u88c5\u5e38\u7528\u7684\u4eba\u548c\u5c0f\u9f20\u7684\u57fa\u56e0\u4fe1\u606f\u6570\u636e\u5e93\nBiocManager::install(c('org.Hs.eg.db', 'org.Mm.eg.db'))\n\n## \u6700\u540e\u5b89\u88c5garnett\ndevtools::install_github(\"cole-trapnell-lab\/garnett\", ref=\"monocle3\")\nlibrary(garnett)\n\ndownload.file(url=\"https:\/\/cole-trapnell-lab.github.io\/garnett\/marker_files\/hsPBMC_markers.txt\",\n              destfile = \"hsPBMC_markers.txt\")\n\ndownload.file(url=\"https:\/\/cf.10xgenomics.com\/samples\/cell-exp\/3.0.2\/5k_pbmc_v3_nextgem\/5k_pbmc_v3_nextgem_filtered_feature_bc_matrix.h5\", \n              destfile = \"pbmc.h5\")\n\ndownload.file(url=\"https:\/\/cole-trapnell-lab.github.io\/garnett\/classifiers\/hsPBMC_20191017.RDS\",\n              destfile = \"hsPBMC.rds\")\ngetwd()\n###\n\n## \u521b\u5efaseurat\u5bf9\u8c61\u5e76\u964d\u7ef4\u805a\u7c7b\n\nload(\"C:\/shangke\/lession11\/scRNA_harmony.rdata\")\npbmc <- scRNA_harmony\ncell=colnames(scRNA_harmony)\npbmc=pbmc[,cell]\npbmc <- SCTransform(pbmc)\npbmc <- RunPCA(pbmc, verbose = F)\nElbowPlot(pbmc)\npc.num=1:15\npbmc <- pbmc %>% RunTSNE(dims=pc.num) %>% RunUMAP(dims=pc.num) %>%\n  FindNeighbors(dims = pc.num) %>% FindClusters(resolution=0.8) \n\n## \u521b\u5efaCDS\u5bf9\u8c61\ndata <- GetAssayData(pbmc, assay = 'RNA', slot = 'counts')\ncell_metadata <- pbmc@meta.data\ngene_annotation <- data.frame(gene_short_name = rownames(data))\n?data.frame\nrownames(gene_annotation) <- rownames(data)\ncds <- new_cell_data_set(data,\n                         cell_metadata = cell_metadata,\n                         gene_metadata = gene_annotation)\n?new_cell_data_set\n#preprocess_cds\u51fd\u6570\u76f8\u5f53\u4e8eseurat\u4e2dNormalizeData+ScaleData+RunPCA\ncds <- preprocess_cds(cds, num_dim = 30)\n\n## \u6f14\u793a\u5229\u7528marker file\u8bad\u7ec3\u5206\u7c7b\u5668\n# \u5bf9marker file\u4e2dmarker\u57fa\u56e0\u8bc4\u5206\nlibrary(org.Hs.eg.db)\nmarker_check <- check_markers(cds, \"hsPBMC_markers.txt\",\n                              db=org.Hs.eg.db,\n                              cds_gene_id_type = \"SYMBOL\",\n                              marker_file_gene_id_type = \"SYMBOL\")\n?check_markers\nplot_markers(marker_check)\n\n# \u4f7f\u7528marker file\u548ccds\u5bf9\u8c61\u8bad\u7ec3\u5206\u7c7b\u5668\npbmc_classifier <- train_cell_classifier(cds = cds,\n                                         marker_file = \"hsPBMC_markers.txt\",\n                                         db=org.Hs.eg.db,\n                                         cds_gene_id_type = \"SYMBOL\",\n                                         num_unknown = 50,\n                                         marker_file_gene_id_type = \"SYMBOL\")\n?train_cell_classifier\nsaveRDS(pbmc_classifier, \"my_classifier.rds\")\n\n#\u67e5\u770b\u5206\u7c7b\u5668\u6700\u540e\u9009\u62e9\u7684\u6839\u8282\u70b9\u57fa\u56e0\uff0c\u6ce8\u610fmarkerfile\u7684\u57fa\u56e0\u90fd\u4f1a\u5728\u5176\u4e2d\nfeature_genes_root <- get_feature_genes(pbmc_classifier, node =\"root\", db= org.Hs.eg.db, \n                                       convert_ids = TRUE)\n?get_feature_genes\nhead(feature_genes_root)\n\n#\u4f7f\u7528garnett\u5b98\u7f51\u8bad\u7ec3\u597d\u7684\u5206\u7c7b\u5668\u9884\u6d4b\u6570\u636e\u3002\ndownload.file(url=\"https:\/\/cole-trapnell-lab.github.io\/garnett\/classifiers\/hsPBMC_20191017.RDS\",\n              destfile = \"hsPBMC.rds\", mode = \"wb\")\n\nhsPBMC <- readRDS(\"hsPBMC.rds\")\n?pData\npData(cds)\npData(cds)$garnett_cluster <- pData(cds)$seurat_clusters\ncds <- classify_cells(cds,\n                      pbmc_classifier, \n                      db = org.Hs.eg.db,\n                      cluster_extend = TRUE,\n                      cds_gene_id_type = \"SYMBOL\")\n\n# \u63d0\u53d6\u5206\u7c7b\u7ed3\u679c\ncds.meta <- subset(pData(cds), select = c(\"cell_type\", \"cluster_ext_type\")) %>% as.data.frame()\n## \u5c06\u7ed3\u679c\u8fd4\u56de\u7ed9seurat\u5bf9\u8c61\n?AddMetaData\npbmc <- AddMetaData(pbmc, metadata = cds.meta)\n\ntsne <- as.data.frame(pbmc@reductions[[\"tsne\"]]@cell.embeddings)\ndata <-  as.data.frame(pData(cds))\ndata_umap <- merge(data,tsne,by=0)\ncolnames(data_umap)\nqplot(tSNE_1, tSNE_2, color = cell_type, data = data_umap) + theme_bw()\n?qplot\n\nqplot(tSNE_1, tSNE_2, color = cluster_ext_type, data = data_umap, labels= T) + theme_bw()\n##\u4e0a\u56fe\u4e2d\u7b2c\u4e00\u4e2a\u56fe\u663e\u793a\u4e86Garnett\u7684\u7ec6\u80de\u7c7b\u578b\u5206\u914d\uff0c\u7b2c\u4e8c\u5f20\u56fe\u663e\u793a\u4e86Garnett\u7684cluster\u7fa4\u6269\u5c55\u7c7b\u578b\u5206\u914d\u3002\n<\/code><\/pre>\n
    1.new_cell_data_set()<\/h3>\n
    new_cell_data_set ( )\uff1a\u4ece\u5934\u521b\u5efaCDS\u5bf9\u8c61\u5e76\u9884\u5904\u7406\u6570\u636e\uff0c\u521b\u5efa\u4e86\u4e00\u4e2a\u5168\u65b0\u7684\u5bf9\u8c61\uff0c\u8fd9\u6837\u5f88\u7e41\u7410\uff0c\u8fd8\u8981\u518d\u505a\u4e00\u6b21\u964d\u7ef4\u805a\u7c7b\u5206\u7fa4\u3002\u56e0\u4e3a\u6211\u4eec\u5bfc\u5165\u4e86seurat\u5bf9\u8c61\u91cc\u7684\u8868\u8fbe\u77e9\u9635\uff0cmeta\u4fe1\u606f\u548cgenelist\uff0c\u6240\u4ee5\u8fd9\u4e2acds\u662f\u6ca1\u6709\u8fdb\u884c\u964d\u7ef4\u805a\u7c7b\u7b49\u64cd\u4f5c\uff0c\u5bfc\u81f4\u540e\u9762\u7684\u62df\u65f6\u5e8f\u5206\u6790\u662f\u505a\u4e0d\u4e86\u7684\uff0c\u5b98\u7f51\u4e0a\u8bf4\u62df\u65f6\u5e8f\u5206\u6790\u662f\u57fa\u4e8e\u4f4e\u7ef4\u5ea6\uff0c\u6240\u4ee5\u5fc5\u987b\u5bf9cds\u8fdb\u884c\u964d\u7ef4\u805a\u7c7b\u5206\u7fa4\u3002<\/p>\n
    new_cell_data_set(expression_data, cell_metadata = NULL, gene_metadata = NULL)\nArguments\nexpression_data \nexpression data matrix for an experiment, can be a sparseMatrix.\n\ncell_metadata   \ndata frame containing attributes of individual cells, where row.names(cell_metadata) = colnames(expression_data).\n\ngene_metadata   \ndata frame containing attributes of features (e.g. genes), where row.names(gene_metadata) = row.names(expression_data).\n<\/code><\/pre>\n
    as.cell_data_set ( ):  \u76f4\u63a5\u8bfb\u5165\u7ecf\u8fc7Seurat\u4e0a\u6e38\u5904\u7406\uff08SCTransform\u3001RunPCA\u3001RunUMAP\u3001FindNeighbors\u3001FindClusters\uff09\u540e\u7684obj\u521b\u5efards\u5bf9\u8c61\uff0c\u4e0d\u7528\u518d\u8fdb\u884c\u9884\u5904\u7406\uff08preprocess_cds\uff09\u548c\u964d\u7ef4\uff08reduce_dimension\uff09\u6b65\u9aa4\u3002<\/p>\n
    2.marker\u6587\u4ef6\u51c6\u5907<\/h3>\n
    \n  1.><\/code>\u5f00\u5934\u7684\u7ec6\u80de\u7c7b\u578b\u884c\uff1b
    \n2.expressed\uff1a<\/code>\u5f00\u5934\u884c\uff0c\u540e\u9762\u8ddf\u5b9a\u4e49\u7ec6\u80de\u7c7b\u578b\u7684marker\u57fa\u56e0\u3002\u57fa\u56e0\u4e4b\u95f4\u4f7f\u7528\uff0c\u5206\u9694\u3002
    \n\u53ef\u9009\u4ee5\u4e0b\u51e0\u884c\u7684\u9644\u52a0\u4fe1\u606f\uff1a
    \n1.not expressed\uff1a<\/code>\u5f00\u5934\u884c\uff0c\u540e\u9762\u8ddf\u5b9a\u4e49\u7ec6\u80de\u7c7b\u578b\u7684\u8d1f\u9009marker\u57fa\u56e0\u3002\u4f8b\u5982CD4+T\u7ec6\u80de\u4e0d\u80fd\u8868\u8fbeCD8\uff0c\u5c31\u53ef\u4ee5\u5199\u5728\u8fd9\u4e00\u884c\u3002
    \n2.\u53ef\u4ee5\u4f7f\u7528expressed above\uff1a<\/code>\u3001expressed below\uff1a<\/code>\u3001expressed between\uff1a<\/code>\u5b9a\u4e49marker\u57fa\u56e0\u7684\u8868\u8fbe\u503c\u8303\u56f4\u3002\u9002\u7528\u4e8e\u4e00\u4e9bmarker\u57fa\u56e0\u662f\u6839\u636ehigh\/Iow\u6765\u533a\u5206\u7ec6\u80de\u7fa4\u7684\u60c5\u51b5\u3002\u6bd4\u5982\u6ce8\u91caLy6 ChighCCR2 highCX3CR1low\u7684\u5355\u6838\u7ec6\u80de\u3002
    \n3.subtype of\uff1a<\/code>\u5f00\u5934\u7684\u5b57\u7b26\u5b9a\u4e49\u7ec6\u80de\u7c7b\u578b\u7684\u7236\u7c7b\uff0c\u5373\u6b64\u7c7b\u7ec6\u80de\u5c5e\u4e8e\u54ea\u79cd\u7ec6\u80de\u7684\u4e9a\u578b\u3002
    \n4.references\uff1a<\/code>\u5f00\u5934\u7684\u5b57\u7b26\u5b9a\u4e49marker\u57fa\u56e0\u7684\u9009\u62e9\u4f9d\u636e\u3002
    \n5.#\u540e\u53ef\u6dfb\u52a0\u6ce8\u91ca<\/p><\/blockquote>\n
    \"\u56fe\u7247[1]-\u5403\u4e86\u5403\u4e86\"<\/p>\n
    3.check_markers()<\/h3>\n
    \u68c0\u67e5\u4e3a\u6807\u8bb0\u6587\u4ef6\u9009\u62e9\u7684\u6807\u8bb0\uff0c\u5e76\u751f\u6210\u6709\u7528\u7684\u7edf\u8ba1\u4fe1\u606f\u8868\u3002\u8be5\u51fd\u6570\u7684\u8f93\u51fa\u7ed3\u679c\u53ef\u8f93\u5165 plot_markers \u751f\u6210\u8bca\u65ad\u56fe\u3002<\/p>\n
    \"\u56fe\u7247[2]-\u5403\u4e86\u5403\u4e86\"<\/p>\n
    \u8bc4\u4f30\u7ed3\u679c\u4f1a\u4ee5\u7ea2\u8272\u5b57\u4f53\u63d0\u793a\u54ea\u4e9bmarker\u57fa\u56e0\u5728\u6570\u636e\u5e93\u4e2d\u627e\u4e0d\u5230\u5bf9\u5e94\u7684Ensembl\u540d\u79f0\uff0c\u4ee5\u53ca\u54ea\u4e9b\u57fa\u56e0\u7684\u7279\u5f02\u6027\u4e0d\u9ad8\uff08\u6807\u6ce8"High overlap with XX cells"\uff09\u3002\u6211\u4eec\u53ef\u4ee5\u6839\u636e\u8bc4\u4f30\u7ed3\u679c\u4f18\u5316narker\u57fa\u56e0\uff0c\u6216\u8005\u6dfb\u52a0\u5176\u4ed6\u4fe1\u606f\u6765\u8f85\u52a9\u533a\u5206\u7ec6\u80de\u7c7b\u578b\u3002<\/p>\n
    4.train_cell_classifier<\/h3>\n
    \u8be5\u51fd\u6570\u4ee5 CDS \u5bf9\u8c61\u548c\u7ec6\u80de\u7c7b\u578b\u5b9a\u4e49\u6587\u4ef6\uff08\u6807\u8bb0\u6587\u4ef6\uff09\u7684\u5f62\u5f0f\u83b7\u53d6\u5355\u7ec6\u80de\u8868\u8fbe\u6570\u636e\uff0c\u5e76\u8bad\u7ec3\u591a\u9879\u5f0f\u5206\u7c7b\u5668\u6765\u5206\u914d\u7ec6\u80de\u7c7b\u578b\u3002\u751f\u6210\u7684 garnett_classifier \u5bf9\u8c61\u53ef\u7528\u4e8e\u5bf9\u540c\u4e00\u6570\u636e\u96c6\u6216\u672a\u6765\u7c7b\u4f3c\u7ec4\u7ec7\/\u6837\u672c\u6570\u636e\u96c6\u4e2d\u7684\u7ec6\u80de\u8fdb\u884c\u5206\u7c7b\u3002<\/p>\n
    5.get_feature_genes<\/h3>\n
    \u4ece\u8bad\u7ec3\u597d\u7684 garnett_classifier \u4e2d\u63d0\u53d6\u88ab\u9009\u4e3a\u7ec6\u80de\u7c7b\u578b\u5206\u7c7b\u7279\u5f81\u7684\u57fa\u56e0\u3002<\/p>\n
    6.pData<\/h3>\n
    \u8bbf\u95ee cds colData \u8868\u7684\u901a\u7528\u65b9\u6cd5<\/p>\n
    \u4e09\u3001Azimuth<\/h1>\n
    \u4f7f\u7528\u94fa\u70b9\u6574\u5408\u7684\u65b9\u6cd5\u5bf9\u5355\u7ec6\u80de\u7c7b\u578b\u8fdb\u884c\u9884\u6d4b\uff0c\u53ef\u4ee5\u7528\u4e8e\u624b\u52a8\u7ec6\u80de\u6ce8\u91ca\u7ed3\u679c\u7684\u53c2\u8003\u3002<\/p>\n
    \u5c06seurat\u5bf9\u8c61\u7684count\u77e9\u9635\u4fdd\u5b58\u4e3ards\u6587\u4ef6\uff0c\u76f4\u63a5\u8f93\u5165Azimuth\u7f51\u7ad9\u8fdb\u884c\u9884\u6d4b\u3002<\/p>\n
    Azimuth\u7ec6\u80de\u9884\u6d4b\u7f51\u5740\uff1ahttps:\/\/satijalab.org\/azimuth\/<\/a><\/p>\n
    \u4e0a\u4f20http:\/\/azimuth.satijalab.org\/app\/azimuth\u7f51\u7ad9\u5728\u7ebf\u5206\u7c7b\uff0c\u5206\u7c7b\u7ed3\u679c\u4e3aazimuth_pred.tsv\u6587\u4ef6<\/p>\n
    pbmc_counts <- pbmc@assays$RNA@counts\nsaveRDS(pbmc_counts, \"pbmc_counts.rds\")\n#\u4e0a\u4f20\u540e\u4e0b\u8f7d\npredictions <- read.delim('azimuth_pred.tsv', row.names = 1)\npbmc <- AddMetaData(pbmc, metadata = predictions)\ncolnames(predictions)\nDimPlot(pbmc, group.by = \"predicted.celltype.l2\", label = T, \n        label.size = 3) +  ggtitle(\"Classified by Azimuth\")+ ggsci::scale_color_igv()\n<\/code><\/pre>\n
    \"\u56fe\u7247[3]-\u5403\u4e86\u5403\u4e86\"<\/p>\n","protected":false},"excerpt":{"rendered":"
    \u4e00\u3001singleR 1. \u603b\u4f53\u4ee3\u7801 #\u7b2c\u4e8c\u79cd\u65b9\u6cd5\u7528SingleR\u9274\u5b9a\u7ec6\u80de\u7c7b\u578b ###\u4e0b\u8f7d\u597d\u6570\u636e\u5e93\u540e\uff0c\u628aref_ […]<\/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":[],"topics":[],"_links":{"self":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/444"}],"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=444"}],"version-history":[{"count":4,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/444\/revisions"}],"predecessor-version":[{"id":500,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/posts\/444\/revisions\/500"}],"wp:attachment":[{"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/media?parent=444"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/categories?post=444"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/tags?post=444"},{"taxonomy":"topics","embeddable":true,"href":"https:\/\/blog.ineuro.net\/index.php\/wp-json\/wp\/v2\/topics?post=444"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}