1. #提取单细胞亚群子集
    2. rm(list = ls())
    3. library(Seurat)
    4. # devtools::install_github('satijalab/seurat-data')
    5. library(SeuratData)
    6. library(ggplot2)
    7. library(patchwork)
    8. library(dplyr)
    9. load(file = 'basic.sce.pbmc.Rdata')
    10. DimPlot(pbmc, 
    11.         reduction = 'umap', 
    12.         label = TRUE, pt.size = 0.5) + NoLegend()
    13. FeaturePlot(pbmc, 
    14.             features = c("CD4"))
    15. sce=pbmc
    17. Idents(sce)
    18. levels(sce)
    19. head(sce@meta.data)
    21. genes_to_check = c('PTPRC', 'CD3D', 'CD3E', 
    22.                    'CD4','IL7R','NKG7','CD8A')
    23. DotPlot(sce, 
    24.         group.by = 'seurat_clusters',
    25.         features = unique(genes_to_check)) + RotatedAxis()
    26. DotPlot(sce, # group.by = 'seurat_clusters',
    27.         features = unique(genes_to_check)) + RotatedAxis()
    29. p1=DimPlot(sce, 
    30.            reduction = 'umap', 
    31.            group.by = 'seurat_clusters',
    32.            label = TRUE, pt.size = 0.5) + NoLegend()
    33. p2=DotPlot(sce, 
    34.            group.by = 'seurat_clusters',
    35.            features = unique(genes_to_check)) + RotatedAxis()
    36. p1+p2
    38. #在R里面取子集方法:逻辑值，坐标，名字
    39. #取子集方法1:
    40. cd4_sce1 = sce[,sce@meta.data$seurat_clusters %in% c(0,2)]
    41. #取子集方法2:
    42. cd4_sce2 = sce[, Idents(sce) %in% c("Naive CD4 T","Memory CD4 T")]
    44. #subset 函数也可以,手写函数也行
    45. cd4_sce1
    46. cd4_sce2
    48. #代码不要变动
    49. sce=cd4_sce1
    50. sce <- NormalizeData(sce, 
    51.                      normalization.method = "LogNormalize", 
    52.                      scale.factor = 1e4) 
    53. sce <- FindVariableFeatures(sce, 
    54.                             selection.method = 'vst', 
    55.                             nfeatures = 2000)
    56. sce <- ScaleData(sce, 
    57.                  vars.to.regress = "percent.mt")
    58. sce <- RunPCA(sce, 
    59.               features = VariableFeatures(object = sce)) 
    61. sce <- FindNeighbors(sce, 
    62.                      dims = 1:10)
    63. sce <- FindClusters(sce, 
    64.                     resolution = 1 )
    66. # Look at cluster IDs of the first 5 cells
    67. head(Idents(sce), 5)
    68. table(sce$seurat_clusters) 
    69. sce <- RunUMAP(sce, 
    70.                dims = 1:10)
    71. DimPlot(sce, 
    72.         reduction = 'umap')
    74. genes_to_check = c('PTPRC', 'CD3D', 'CD3E', 'FOXP3',
    75.                    'CD4','IL7R','NKG7','CD8A')
    76. DotPlot(sce, 
    77.         group.by = 'seurat_clusters',
    78.         features = unique(genes_to_check)) + RotatedAxis()
    80. # 亚群水平 
    81. p1=DimPlot(sce, 
    82.            reduction = 'umap', 
    83.            group.by = 'seurat_clusters',
    84.            label = TRUE, pt.size = 0.5) + NoLegend()
    85. p2=DotPlot(sce, 
    86.            group.by = 'seurat_clusters',
    87.            features = unique(genes_to_check)) + RotatedAxis()
    88. p1+p2
    90. save(sce,file = 'sce.cd4.subset.Rdata')
    92. load(file = 'sce.cd4.subset.Rdata')
    94. #先执行不同resolution 下的分群
    95. library(Seurat)
    96. library(clustree)
    97. sce <- FindClusters(
    98.   object = sce,
    99.   resolution = c(seq(.1,1.6,.2))
    100. )
    101. clustree(sce@meta.data, prefix = "RNA_snn_res.")

    image.png
    image.png