################某基因与多基因间相关性 表达

rm(list = ls())

setwd("/data3/ff/data/xena_TCGA/TPM/gene.lnc")

#####读入表达文件                   

library(data.table)

PAAD<-fread("TCGA-PAAD.htseq-TPM.Gene.lnc.txt",data.table = F)

row.names(PAAD)<-PAAD$genenames

###########提取所需基因

gene<-read.table("/data3/ff/data/TF/AL5.TF.txt",header = T)

gene<-gene$TF

#####在表达文件中提取所需基因

expr<-PAAD[gene,]

#转置

expr<-t(expr)

expr<-as.data.frame(expr)

#删除第一行

expr<-expr[-1,]

#使用Hmisc 包，计算矩阵相关系数及其对应的显著性水平

library(Hmisc)

#写函数来提取矩阵相关及其P值

CorMatrix <- function(cor,p) {

                              ut <- upper.tri(cor)

                              data.frame(row = rownames(cor)[row(cor)[ut]] ,

                              column = rownames(cor)[col(cor)[ut]],

                              cor =(cor)[ut],

                              p = p[ut] )

}

res <- rcorr(as.matrix(expr), type=c("pearson"))

?rcorr #查看函数说明    rcorr(x, y, type=c("pearson","spearman"))

library(dplyr)

results <- CorMatrix (res$r, res$P) %>% as.data.frame() %>%

        filter(row == "AL5" & cor <0)

write_csv(results,"AL5.negative.TF.csv",col_names = T)

#####################两基因间相关性图

rm(list = ls())

setwd("/data3/ff/data/xena_TCGA/TPM/gene.lnc")

PAAD<-fread("TCGA-PAAD.htseq-TPM.Gene.lnc.txt",data.table = F)

row.names(PAAD)<-PAAD$genenames

PAAD<-PAAD[,-1]

##############################################################---LA.HE1相关性

LA.HE1<-PAAD[c("LA","HE1"),]

#转置

LA.HE1<-t(LA.HE1)

LA.HE1<-as.data.frame(LA.HE1)

#################################画图

library(ggplot2)

library(ggpmisc)

library(ggpubr)

#使用ggpubr包stat_cor，将相关系数和显著性水平加入

my.formula <- y ~ x

ggplot(data=LA.HE1,aes(x=LA,y=HE1)) +

  scale_x_log10() + scale_y_log10() +

  labs(title="TCGA PAAD", x="LA expression", y="HE1 expression")+

  theme_bw()+#移除背景

  theme(panel.grid.major=element_blank(),panel.grid.minor=element_blank())+#移除网格线

  theme(plot.title = element_text(hjust = 0.5))+ #设置标题居中

  geom_point(size=2,alpha=1,color="black") +

  #stat_poly_eq(formula = my.formula,

  #aes(label = paste(#..eq.label..,#展示方程

  #..rr.label..,#展示r值

  # sep = "~~~")),

  #parse = TRUE) +

  stat_cor(method = "pearson",#将相关系数和显著性水平加入

          #label.x =-2, label.y = -2,

          color='black',p.accuracy = 0.001)+

  geom_smooth(method="lm",color="red",alpha=.7,size=1.2,se=FALSE, formula = my.formula)+

  scale_y_continuous(breaks = c(0,1,2,3,4,5,6,7,8,9,10))+

  scale_x_continuous(breaks = c(0,1,2,3,4,5,6,7,8,9,10));

ggsave("LA.HE1.cor.pdf",width = 10, height =8,dpi = 300)  #保存成pdf