其他
R可视化——进化树+柱状堆积图组合！
设置工作环境并加载所需R包
rm(list=ls())#clear Global Environmentsetwd('D:\\桌面\\test')#设置工作路径
#加载R包library (reshape2)library(ggplot2) # Create Elegant Data Visualisations Using the Grammar of Graphicslibrary(plyr) # Tools for Splitting, Applying and Combining Datalibrary(ggtree) # an R package for visualization of tree and annotation datalibrary(vegan) # Community Ecology Packagelibrary(ape) # Analyses of Phylogenetics and Evolutionlibrary( RColorBrewer) # not installed on this machinelibrary(aplot) # Decorate a 'ggplot' with Associated Information加载数据并进行预处理
#读取数据df1 <- read.table(file="Genus.txt",sep="\t",header=T,check.names=FALSE)#查看前6行head(df1)##利用循环处理具有重复的数据#初始化data<-aggregate(E ~ Tax,data=df1,sum)#重命名colnames(data)[2]<-"example"for (i in colnames(df1)[2:length(colnames(df1))]){ #计算每列的和 data1<-aggregate(df1[,i]~Tax,data=df1,sum) colnames(data1)[2]<-i #合并列 data<-merge(data,data1,by="Tax") }df2<-data[,-2]rownames(df2)=df2$Tax#修改行名df3=df2[,-1]#删除多的列
#计算物种总丰度并降序排列df3$rowsum <- apply(df3,1,sum)df4 <- df3[order (df3$rowsum,decreasing=TRUE),]df5 = df4[,-6]#删除求和列#求物种相对丰度df6 <- apply(df5,2,function(x) x/sum(x))#由于之间已经按照每行的和进行过升序排列，所以可以直接取前10行df7 <- df6[1:10,]df8 <- 1-apply(df7, 2, sum) #计算剩下物种的总丰度#合并数据df9 <- rbind(df7,df8)row.names(df9)[11]="Others"#导出数据write.table (df9, file ="genus_x.csv",sep =",", quote =FALSE)#变量格式转换,宽数据转化为长数据,方便后续作图df_genus <- melt(df9)names(df_genus)[1:2] <- c("Taxonomy","sample") #修改列名柱状堆积图的绘制
#颜色col <-colorRampPalette(brewer.pal(12,"Paired"))(11)##柱状堆积图绘制绘图p1<-ggplot(df_genus, aes( x = sample,y=100 * value,fill = Taxonomy))+#geom_col和geom_bar这两条命令都可以绘制堆叠柱形图 geom_col(position = 'stack',width = 0.8)+ #geom_bar(position = "stack", stat = "identity", width = 0.6) scale_y_continuous(expand = c(0,0))+# 调整y轴属性，使柱子与X轴坐标接触 labs(x=NULL,y="Relative Abundance(%)",#设置X轴和Y轴的名称以及添加标题 fill="Taxonomy")+ guides(fill=guide_legend(keywidth = 1, keyheight = 1)) + coord_flip()+ theme(panel.grid = element_blank(), panel.background = element_blank())+ scale_fill_manual(values = col)p1进化树的绘制
data2 <- df3[-6]#计算距离矩阵df_dist <- vegdist(t(data2),method = 'bray')#使用bray curtis方法计算距离矩阵####进行层次聚类,可选择方法有single、complete、median、mcquitty、average 、centroid 、ward df_hc <- hclust(df_dist,method="average")#使用类平均法进行聚类#绘图plot(as.dendrogram(df_hc),type="rectangle",horiz=T)#实现将垂直的聚类树变成水平聚类树，绘图# 将聚类结果转成系统发育格式df_tree <- as.phylo(df_hc)# 对树分组gro <- list(group1=c("A","E"), group2=c("C","D"), group3=c("B"))# 将分组信息和进化树组合到一起tree<-groupOTU(df_tree,gro)# ggtree绘图p2 <- ggtree(tree,size=1.2)+ geom_tiplab(aes(color=group),size=5,align = F, offset = 0.008,show.legend = F)+ geom_tippoint(aes(shape=group,color=group), show.legend = T, size=4)p2柱状堆积图与进化树的合并
使用aplot包将柱状堆积图与进化树进行拼接匹配合并，这里要注意查看柱状堆积图的顺序与进化树的顺序要匹配：
p1%>%insert_left(p2,width = 0.3)美化
p1 <- p1+theme(axis.text.y = element_text(color = "black",size=12))col2 <-colorRampPalette(brewer.pal(9,"Set1"))(3)p3 <- ggtree(tree,size=1)+ geom_tippoint(aes(shape=group,color=group), show.legend = T, size=4)+ scale_color_manual(values = col2)p <- p1%>%insert_left(p3,width = 0.3)p爱我请给我好看!