相关性热图、圈图、弦图（笔记）

#感兴趣基因/样本的相关性图----
rm(list = ls()) 
load(file = "step1output.Rdata")
load(file = "step4output.Rdata")
load(file = 'step2output.Rdata')
library(dplyr)
library(ggplot2)
library(pheatmap)
library(corrplot)
dat = deg[!duplicated(deg$symbol),]
exp = exp[dat$probe_id,]
rownames(exp) = dat$symbol
g = c("TIMM50", "GOLGA2", "CLIP1", "DDX6", "CTSZ", "P2RY12", "LTF", 
      "HBD", "F3", "ZFP36L2")
a = exp[g,]
M = cor(t(exp[g,])) #cor()函数可以对矩阵计算相关性（谁在列的位置就计算谁的相关性）
#画基因之间的相关性，cor函数后面的矩阵exp[g,]要以基因为列名（转置一下）
#画样本之间的相关性，cor函数后面的矩阵exp[g,]要以样本为列名（不要转置）

#相关性热图
pheatmap(M) 

#相关性圆圈图
library(paletteer)
my_color = rev(paletteer_d("RColorBrewer::RdYlBu"))
my_color
my_color = colorRampPalette(my_color)(10) #整成10个色阶
# library(RColorBrewer)
# my_color=rev(brewer.pal(10, "RdYlBu"))
corrplot(M, type="upper", 
         method="pie",
         order="hclust", 
         col=my_color,
         tl.col="black", 
         #tl.pos = "d",
         tl.srt=45         
        )

#corrplot不能赋值，和其他R包拼图就比较费劲
library(cowplot) #比patchwork包更强的拼图包
cor_plot <- recordPlot() #cowplot包里的recordPlot函数可以把这个相关性图抠下来强行赋值

# 拼图（相关性图属于另外一个拼图体系）
#load("pca_plot.Rdata")
pdf("cor_plot.pdf", width = 10, height = 10)
plot_grid(cor_plot)
#plot_grid(pca_plot,cor_plot) #cowplot包里的plot_grid()函数允许跨绘图体系拼图
dev.off()

## 圆圈视图
M = cor(t(a))
p.mat <- cor.mtest(t(a))$p
library(paletteer)
my_color = rev(paletteer_d("RColorBrewer::RdYlBu")[-1])
my_color = colorRampPalette(my_color)(10)
png("plot/corrplot.png",height = 1000,width = 1000)
corrplot(M, type="upper", 
         order="hclust", 
         col = my_color,
         p.mat = p.mat, 
         sig.level = 0.01, 
         insig = "blank",
         tl.col = "black",
         tl.srt=45)
dev.off()

## 相关性弦图
library(circlize)
library(tidyr)
library(tibble)
library(ComplexHeatmap)

mat = M
df = mat %>%
  as.data.frame() %>%
  rownames_to_column(var = "from") %>%
  pivot_longer(names_to  = "to",
               values_to  = "value",
               cols = 2:(nrow(mat)+1) #有10个基因，这里就是11
               ) 
df = df[df$from != df$to,]
df = df[!duplicated(df$value),]

#自定义边的颜色
library(RColorBrewer)
col_sample = c(brewer.pal(n = 12,name = "Set3")
               #brewer.pal(n = 8,name = "Set2")
)
col_sample
border_color <- sample(col_sample,nrow(mat)) 
#border_color = c("#66C2A5", "#FC8D62", "#8DA0CB", "#E78AC3") 

#根据相关性大小展示连线的颜色范围：红 白 绿
col = colorRamp2(c(-1, 0, 1), c('green', 'white', 'red'), transparency = 0.5)
# range(mat)
# col <- colorRamp2(range(mat), c("#FA0A0A", "#FFFFFF")) 

# ComplexHeatmapt添加图例
lgd = Legend(col_fun = col, title = "foo")
ComplexHeatmap:::width(lgd)
ComplexHeatmap:::height(lgd)

pushViewport(viewport(width = 0.9, height = 0.9))
grid.rect()  # border
draw(lgd, x = unit(1, "cm"), y = unit(1, "cm"), just = c("left", "bottom"))
# draw(lgd, x = unit(0.5, "npc"), y = unit(0.5, "npc"))
# draw(lgd, x = unit(1, "npc"), y = unit(1, "npc"), just = c("right", "top"))
popViewport()


chordDiagram(x = df, 
             directional = 0, #表示线条的方向，0代表没有方向，1代表正向，-1代表反向，2代表双向
             grid.col = border_color, #圈的颜色
             col = col, #弦的颜色
             annotationTrack = c('grid', 'name', 'axis'), #绘制外周圆弧区，显示名称和刻度轴
             annotationTrackHeight = c(0.05, 0.1),#外面一圈的宽度
             transparency = 0.25#线条的透明度
)