震撼!DeepSeek丝滑完成单细胞数据初步分析~~

BacQ · 发表于 5 天前

顶刊易上易！60例食管癌单细胞测序为何出收顶刊，而是行步于NC？

2025-06-06

<img />

<img />

DeepSeek！ggplot2丝滑画造丛林图！

2025-06-07

震动!DeepSeek丝滑完毕单细胞数据开端阐发~~w2.jpg

<img />

<img />

完整DeepSeek，代码丝滑！GSE112973，多分组百般原，数据洗濯战差别阐发，水山图！

2025-06-05

震动!DeepSeek丝滑完毕单细胞数据开端阐发~~w3.jpg

<img />

<img />

DeepSeek太赞了！假设导师让您师长教师疑，无妨先尝尝那份ggplot2 可望化代码！

2025-06-06

震动!DeepSeek丝滑完毕单细胞数据开端阐发~~w4.jpg

<img />

<img />
按照附件文件，并按照以上疑息，供给GSE160269食管癌CD45neg单细胞数据读与，Seurat数据工具建立，细胞分群战细胞通信的代码（R剧本)，所需的UMI战cells文献已经下载正在当地，读与数据用fread()函数读与。按照文件，CD45neg细胞包罗：上皮细胞、成纤维细胞、内乱皮细胞、周细胞战成纤维网状细胞。数据阐发步调：1. 读与UMI矩阵，使用fread函数，留神文献是gz收缩的，fread能够间接读与。2. 读与细胞正文文献，那个文献该当包罗每一个细胞的范例疑息。3.创立 Seurat工具，使用UMI矩阵，并截至根本的量控战尺度化。4.依据正文文献增加细胞范例疑息。5.中止尺度阐发：回一化、找下变基果、落维（PCA）、散类（比方使用Louvain算法）战UMAP落维可望化。6.中止细胞通信阐发（能够使用CellChat包）。######################gse160269,食管癌数据开掘########################getwd()setwd("/media/desk16/iyun706/GSE160269")Sys.setenv("VROOM_CONNECTION_SIZE"=99999999)
# 减载须要的R包library(Seurat)library(data.table)library(tidyverse)library(patchwork)library(CellChat)library(Matrix)
### 1. 数据读与取Seurat工具建立 #### 读与UMI矩阵（留神：文献路子需按照理论情况改正）cd45neg_umis <- fread("GSE160269_CD45neg_UMIs.txt.gz", header = F)genes <- cd45neg_umis$V1  # 提炼基果名cd45neg_umis$V1 <- NULL # 移除基果名列
# 变换为稠密矩阵并创立Seurat工具cd45neg_matrix <- as(as.matrix(cd45neg_umis), "dgCMatrix")rownames(cd45neg_matrix) <- genesseurat_obj <- CreateSeuratObject(counts = cd45neg_matrix, project = "ESCC_CD45neg")
# 增加细胞正文（需下载GSE160269_CD45neg_cells.txt.gz）cell_annotations <- fread("GSE160269_CD45neg_cells.txt.gz")seurat_obj <- AddMetaData(seurat_obj, metadata = cell_annotations$annotated_type, col.name = "cell_type")
### 2. 数据预处置取品质掌握 #### 计较线粒体基果比率seurat_obj[["percent.mt"]] <- PercentageFeatureSet(seurat_obj, pattern = "^MT-")
# 量控过滤（按照文件尺度）seurat_obj <- subset(seurat_obj,                   subset = nFeature_RNA > 500 &                      percent.mt < 20)
# 数据尺度化取特性挑选seurat_obj <- NormalizeData(seurat_obj)seurat_obj <- FindVariableFeatures(seurat_obj, selection.method = "vst")seurat_obj <- ScaleData(seurat_obj, vars.to.regress = "percent.mt")
### 3. 落维取散类阐发 #### 线性落维seurat_obj <- RunPCA(seurat_obj, npcs = 50)
# 非线性落维取散类seurat_obj <- RunUMAP(seurat_obj, dims = 1:30)seurat_obj <- FindNeighbors(seurat_obj, dims = 1:30)seurat_obj <- FindClusters(seurat_obj, resolution = 0.5)
# 可望化p1 <- DimPlot(seurat_obj, reduction = "umap", group.by = "cell_type",             label = TRUE) + ggtitle("文件正文细胞范例")p2 <- DimPlot(seurat_obj, reduction = "umap", label = TRUE) + ggtitle("无监视散类成果")p1 + p2
### 4. 细胞范例审定（鉴于文件标识表记标帜基果）###celltype_markers <- list(  Epithelial = c("EPCAM", "KRT5", "KRT14"),  Fibroblasts = c("COL1A1", "DCN", "COL3A1"),  Endothelial = c("VWF", "PECAM1", "CDH5"),  Pericytes = c("RGS5", "MCAM", "ACTA2"),  FRC = c("CCL21", "PDPN"))
#标志基果可望化DotPlot(seurat_obj, features = celltype_markers) + RotatedAxis() + scale_color_gradient2(low = "blue", mid = "gray", high = "red")
### 5. 细胞通信阐发（CellChat）#### 提炼数据并创立CellChat工具data.input <- GetAssayData(seurat_obj, slot = "data", assay = "RNA")meta.data <- seurat_obj@meta.datacellchat <- createCellChat(object = data.input, meta = meta.data, group.by = "cell_type")
# 减载配体-受体数据库CellChatDB <- CellChatDB.humancellchat@DB <- CellChatDB
# 预处置并揣度通信收集cellchat <- subsetData(cellchat)cellchat <- identifyOverExpressedGenes(cellchat)cellchat <- identifyOverExpressedInteractions(cellchat)cellchat <- computeCo妹妹unProb(cellchat)cellchat <- filterCo妹妹unication(cellchat, min.cells = 10)
# 调整通路阐发cellchat <- computeCo妹妹unProbPathway(cellchat)cellchat <- aggregateNet(cellchat)
### 6. 细胞通信可望化 #### 1. 部分阐发，可望化细胞通信收集groupSize <- as.numeric(table(cellchat@idents))par(mfrow = c(1,2), xpd=TRUE)netVisual_circle(cellchat@net$count, vertex.weight = groupSize,                weight.scale = T, label.edge= F,                title.name = "Number of interactions")netVisual_circle(cellchat@net$weight, vertex.weight = groupSize,                weight.scale = T, label.edge= F,                title.name = "Interaction weights/strength")
# 可望化一定旌旗灯号通路pathways.show <- c("PDGF", "VEGF", "MIF") #依据文件挑选主要通路for (pathway in pathways.show) {  netVisual_aggregate(cellchat, signaling = pathway, layout = "circle")}
# 2. 配体-受体互做深度阐发（散焦枢纽通路）# 重心阐发文档中考证的通路（图4/5）pathways <- c("PDGF", "MHC-I", "MHC-II", "CCL", "PD-L1") cellchat@DB <- subsetDB(CellChatDB.human, search = pathways)
# 3. 血管天生取CAF互做（图5）## 3.1 PDGFB-PDGFR互做考证（图5i-k）pairLR.PDGF <- extractEnrichedLR(cellchat, signaling = "PDGF")netVisual_individual(cellchat,                   signaling = "PDGF",                   pairLR.use = pairLR.PDGF[1, ],                   layout = "circle")
## 3.2 内乱皮细胞-周细胞通信冷图（图5g）netVisual_heatmap(cellchat,                signaling = "ANGPT",                color.heatmap = "Reds",                cluster.cols = T                )
# 4. 上皮法式化特性联系关系阐发（图6）## 粘膜法式化细胞通信特性cellchat <- identifyOverExpressedGenes(cellchat)cellchat <- netAnalysis_computeCentrality(cellchat)netAnalysis_signalingRole_heatmap(cellchat,                                  pattern = "outgoing",                               font.size = 5,                               color.heatmap = "Reds")
# 树立图形输出尺微暇（枢纽调解）options(repr.plot.width = 12, repr.plot.height = 10)  # 年夜幅增加绘布尺微暇
# 劣化冷图画造参数netAnalysis_signalingRole_heatmap(cellchat,                                  pattern = "outgoing",                               font.size = 8,       # 删年夜字体                               color.heatmap = "Reds",                               height = 20,          # 冷图下度(厘米)                               width = 18)          # 冷图严度(厘米)
震动!DeepSeek丝滑完毕单细胞数据开端阐发~~w5.jpg