快速上手！热图+表达趋势折线图+通路富集结果组合可视化（含代码） |数据挖掘与分析

● 

ClusterGVis是一个单细胞数据处理可视化的R包，由中国药科大学Jun Zhang博士开发的系列可视化工具包之一，可以同时绘制聚类+分组表达趋势折线图+功能注释的组合图，通过一张热图可以了解差异基因可以划分成几个cluster,每个cluster的表达随着时间是如何变化，以及这些cluster变化的基因通过GO或者KEGG功能注释了解其功能。

如下根据其github以及微信教程简单整理一下其用法。

1. R包（以及依赖包）的安装

# install.packages("devtools")

devtools::install_github("junjunlab/ClusterGVis")

BiocManager::install(“org.Hs.eg.db”)

install.packages(“ggplot2”)

2. 导入需要的R包

library(ClusterGVis)

library(org.Hs.eg.db)

library(Seurat)

library(dplyr)

3. 代码展示

#导入单细胞数据

pbmc <- readRDS("./pbmc3k_final.rds")

#细胞注释

new.cluster.ids <- c("Naive CD4 T", "CD14+ Mono", "Memory CD4 T", "B", "CD8 T")

names(new.cluster.ids) <- levels(pbmc)

pbmc <- RenameIdents(pbmc, new.cluster.ids)

#寻找标记基因

pbmc.markers.all <- Seurat::FindAllMarkers(pbmc,

                                     only.pos = TRUE,

                                     min.pct = 0.25,

                                     logfc.threshold = 0.25)

#选取top10的标记基因

pbmc.markers <- pbmc.markers.all %>%

  dplyr::group_by(cluster) %>%

  dplyr::top_n(n = 10, wt = avg_log2FC)

#利用prepareDataFromscRNA函数准备数据，showAverage 参数设为 TRUE 则表示对基因细胞亚群一样的细胞取均值进行绘图,否则就是所有细胞进行绘图,默认使用seurat对象的 RNA assay的data数据。

st.data <- prepareDataFromscRNA(object = pbmc,

                                diffData = pbmc.markers,

                                showAverage = TRUE)

#对每个Cluster进行GO富集分析（或KEGG）

enrich <- enrichCluster(object = st.data,

                        OrgDb = org.Hs.eg.db,

                        type = "BP",

                        organism = "hsa",

                        pvalueCutoff = 0.5,

                        topn = 5,

                        seed = 5201314)

#挑选需要展示的特征基因

markGenes = unique(pbmc.markers$gene)[sample(1:length(unique(pbmc.markers$gene)),40,

                                             replace = F)]

#绘制表达折线图

visCluster(object = st.data, plot.type = "line")

#绘制热图

pdf('sc1.pdf',height = 10,width = 6,onefile = F)

visCluster(object = st.data,

           plot.type = "heatmap",

           column_names_rot = 45,

           markGenes = markGenes,

           cluster.order = c(1:9))

dev.off()

# heatmap plot

pdf('sc3.pdf',height = 10,width = 6,onefile = F)

visCluster(object = st.data,

           plot.type = "heatmap",

           column_names_rot = 45,

           markGenes = c("CD3D","CD3D_1","CD3D_2"),

           cluster.order = c(1:9))

dev.off()

# 绘制所有细胞

# no average cells

pbmc.markers1 <- pbmc.markers.all %>%

  dplyr::group_by(cluster) %>%

  dplyr::top_n(n = 6, wt = avg_log2FC)

# retain duplicate diff gene in multiple clusters

st.data <- prepareDataFromscRNA(object = pbmc,

                                diffData = pbmc.markers1,

                                showAverage = FALSE)

# heatmap plot

pdf('sc4.pdf',height = 10,width = 8,onefile = F)

visCluster(object = st.data,

           plot.type = "heatmap",

           markGenes = unique(pbmc.markers1$gene),

           column_title_rot = 45,

           cluster.order = 1:9,

           show_column_names = F)

dev.off()

# 修改亚群顺序和修改注释颜色

# change celltype order and color

pdf('sc5.pdf',height = 10,width = 8,onefile = F)

visCluster(object = st.data,

           plot.type = "heatmap",

           markGenes = unique(pbmc.markers1$gene),

           column_title_rot = 45,

           cluster.order = 1:9,

           show_column_names = F,

           sample.cell.order = rev(new.cluster.ids),

           sample.col = jjAnno::useMyCol("paired",n = 9))

dev.off()

# 添加富集注释

# add GO annotation

pdf('sc6.pdf',height = 12,width = 16,onefile = F)

visCluster(object = st.data,

           plot.type = "both",

           column_title_rot = 45,

           markGenes = unique(pbmc.markers1$gene),

           markGenes.side = "left",

           # annoTerm.data = enrich,

           show_column_names = F,

           line.side = "left",

           cluster.order = c(1:9),

           add.bar = T)

dev.off()

ClusterGVis包对于转录组数据的可视化效果非常好，用一张热图就可以展示基因的表达模式，聚类和GO功能注释,大家感兴趣的可以去github(https://github.com/junjunlab/ClusterGVis)深入学习。

相关阅读

点击下方图片进入云平台资料汇总：

所见即所得，绘图高规格联川云平台，让科研更自由