## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 加载包
library(readxl)
library(dplyr)
##
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
library(Hmisc)
## Loading required package: lattice
## Loading required package: survival
## Loading required package: Formula
## Loading required package: ggplot2
## Warning: package 'ggplot2' was built under R version 3.3.2
##
## Attaching package: 'Hmisc'
## The following objects are masked from 'package:dplyr':
##
## combine, src, summarize
## The following objects are masked from 'package:base':
##
## format.pval, round.POSIXt, trunc.POSIXt, units
theme_set(theme_bw(base_family = "STKaiti"))
## 读取数据
ansdata <- read_excel("Array_data_nobackground_normalised.xlsx")
summary(ansdata)
## Factor Factor_group Array data
## Length:6400 Length:6400 Min. : 0.76
## Class :character Class :character 1st Qu.: 80.23
## Mode :character Mode :character Median : 143.06
## Mean : 1487.96
## 3rd Qu.: 316.36
## Max. :59801.26
##
## Patient Id Collection_Date Illness_Day Outcome
## Length:6400 Length:6400 Min. : 3.00 Length:6400
## Class :character Class :character 1st Qu.:11.00 Class :character
## Mode :character Mode :character Median :15.00 Mode :character
## Mean :17.89
## 3rd Qu.:23.00
## Max. :68.00
## NA's :400
## Age Sex Disease Phase Sampling Categry
## Min. : 6.00 Length:6400 Min. :1.0 Length:6400
## 1st Qu.:41.00 Class :character 1st Qu.:2.0 Class :character
## Median :55.00 Mode :character Median :2.0 Mode :character
## Mean :53.67 Mean :2.2
## 3rd Qu.:67.00 3rd Qu.:3.0
## Max. :82.00 Max. :4.0
## NA's :400 NA's :400
## 处理缺失值
colnames(ansdata) <- c("factor","factorgroup","arraydata","paintID",
"collectiondate","illnessday","outcom","age","sex",
"diseaPH","SampleCategry")
summary(ansdata)
## factor factorgroup arraydata
## Length:6400 Length:6400 Min. : 0.76
## Class :character Class :character 1st Qu.: 80.23
## Mode :character Mode :character Median : 143.06
## Mean : 1487.96
## 3rd Qu.: 316.36
## Max. :59801.26
##
## paintID collectiondate illnessday outcom
## Length:6400 Length:6400 Min. : 3.00 Length:6400
## Class :character Class :character 1st Qu.:11.00 Class :character
## Mode :character Mode :character Median :15.00 Mode :character
## Mean :17.89
## 3rd Qu.:23.00
## Max. :68.00
## NA's :400
## age sex diseaPH SampleCategry
## Min. : 6.00 Length:6400 Min. :1.0 Length:6400
## 1st Qu.:41.00 Class :character 1st Qu.:2.0 Class :character
## Median :55.00 Mode :character Median :2.0 Mode :character
## Mean :53.67 Mean :2.2
## 3rd Qu.:67.00 3rd Qu.:3.0
## Max. :82.00 Max. :4.0
## NA's :400 NA's :400
## 对数据进行分析
usedata2 <- ansdata%>%
group_by(factor,factorgroup,sex,outcom)
usedata2$outSC <- paste(usedata2$outcom,usedata2$SampleCategry,sep = "-")
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
# ## -----------该段不实主要的，可用下面的函数定义-------------------------------
# ## "Healthy-First"数据表中细胞自建的相关性
# ## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
# ## 准备数据
# corrdata <- usedata2[usedata2$outSC == "Healthy-First",c(1,2,3,12)]
# corrdata2 <- corrdata[c("factor","arraydata")]
# corrdata2$factor <- factor(corrdata2$factor)
#
# ## 查看因子的个数
# table(corrdata2$factor)
# n <- length(table(corrdata2$factor))
# n
# ## 变换数据
# corrdata3 <- as.data.frame(split(corrdata2$arraydata,corrdata2$factor))
# # aa <- cor(corrdata3,method = "pearson")
# # cor.test(corrdata3)
# ## 计算相关系数
# sper <- rcorr(as.matrix(corrdata3),type = "pearson")
# ## 相关系数矩阵
# sper$r
# ## 对应的p值矩阵
# sper$P
#
# ## 保存数据
# write.csv(sper$r,"相关系数.csv")
# write.csv(sper$P,"相关系数P值.csv")
#
# ## ------------------------------------------------------------------------
# ## 相关系数矩阵
# rellation <- as.data.frame(sper$r)
# rellp <- as.data.frame(sper$P)
# ## 找到每个因子的PR值
# ## 生成相应的基因组合
# hang <- matrix(rownames(rellation), nrow = n, ncol = n,byrow = FALSE)
# lie <- matrix(colnames(rellation),nrow = n,ncol = n,byrow = TRUE)
# ## 对相应的数据取下三角形的数值
# lowrel <- rellation[lower.tri(rellation)]
# lowp <- rellp[lower.tri(rellp)]
# lowhang <- hang[lower.tri(hang)]
# lowlie <- lie[lower.tri(lie)]
# ## 生成数据表
# healthdata <- data.frame(data = matrix(nrow = length(as.vector(lowhang)),ncol = 5))
# names(healthdata) = c("factor1","factor2","relation","relP","factorgroup")
#
# healthdata$factor1 <- as.vector(lowhang)
# healthdata$factor2 <- as.vector(lowlie)
# healthdata$relation <- as.vector(lowrel)
# healthdata$relP <- as.vector(lowp)
# healthdata$factorgroup <- rep("Healthy-First",dim(healthdata)[1])
# ## 保存healthdata
# # write.csv(healthdata,"health的结果.csv",row.names = FALSE)
#
#
# ## 找出符合要求的数据
# healthfuhe <- healthdata[which((healthdata$relation >=0.8 | healthdata$relation <=-0.8) &
# healthdata$relation !=1 & healthdata$relP <= 0.05),]
#
#
# # write.csv(healthfuhe,"health符合要求的数据.csv",row.names = FALSE)
# ## -----------------------------------------------------------------------
# ## -----------该段不实主要的，可用下面的函数定义-------------------------------
## 编写一个求解需要结果的函数
use5data <- function(data,minre1 = 0.8,minre2 = -0.8,maxp = 0.05,dataname = "Healthy-First"){
## 该函数用来计算相关系数，并且输出符合要求的组合
# data <- corrdata2 ## 计算需要的数据，第一列为因子分类数据，第二列为相应的浓度数据
# # data : 数据，一列为基因，另一列为相应的浓度
# minre1 = 0.8 ## 最小的正相关系数
# minre2 = -0.8 ## 最大的富相关系数
# maxp = 0.05 ## 最大的相关系数显著性值
# dataname = "Healthy-First" ## 该数据表的类型,该数据表的名字
library(Hmisc)
## 查看因子的个数
corrdata2 <- data
names(corrdata2) <- c("factor","arraydata")
## 查看因子的个数
n <- length(table(corrdata2$factor))
## 变换数据
corrdata3 <- as.data.frame(split(corrdata2$arraydata,corrdata2$factor))
## 计算相关系数
sper <- rcorr(as.matrix(corrdata3),type = "pearson")
## ------------------------------------------------------------------------
## 相关系数矩阵
rellation <- as.data.frame(sper$r)
rellp <- as.data.frame(sper$P)
## 找到每个因子的PR值
## 生成相应的基因组合
hang <- matrix(rownames(rellation), nrow = n, ncol = n,byrow = FALSE)
lie <- matrix(colnames(rellation),nrow = n,ncol = n,byrow = TRUE)
## 对相应的数据取下三角形的数值
lowrel <- rellation[lower.tri(rellation)]
lowp <- rellp[lower.tri(rellp)]
lowhang <- hang[lower.tri(hang)]
lowlie <- lie[lower.tri(lie)]
## 生成数据表
healthdata <- data.frame(data = matrix(nrow = length(as.vector(lowhang)),ncol = 5))
names(healthdata) = c("factor1","factor2","relation","relP","factorgroup")
healthdata$factor1 <- as.vector(lowhang)
healthdata$factor2 <- as.vector(lowlie)
healthdata$relation <- as.vector(lowrel)
healthdata$relP <- as.vector(lowp)
healthdata$factorgroup <- rep(dataname,dim(healthdata)[1])
## 找出符合要求的数据
healthfuhe <- healthdata[which((healthdata$relation >=minre1 | healthdata$relation <=minre2) &
healthdata$relation !=1 & healthdata$relP <= maxp),]
## 保存数据
res <- list()
res$reldata <- healthdata
res$usereldata <- healthfuhe
return(res)
}
## "Healthy-First"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
corrdata <- usedata2[usedata2$outSC == "Healthy-First",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Healthy-First")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"health的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"health符合要求的数据.csv",row.names = FALSE)
## "Fatal-All other"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Fatal-All other",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Fatal-All other")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Fatal-All other的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Fatal-All other符合要求的数据.csv",row.names = FALSE)
## "Fatal-First"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Fatal-First",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Fatal-First")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Fatal-First的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Fatal-First符合要求的数据.csv",row.names = FALSE)
## "Fatal-Last"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Fatal-Last",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Fatal-Last")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Fatal-Last的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Fatal-Last符合要求的数据.csv",row.names = FALSE)
## "Mild-All other"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Mild-All other",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Mild-All other")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Mild-All other的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Mild-All other符合要求的数据.csv",row.names = FALSE)
## "Mild-First"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Mild-First",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Mild-First")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Mild-First的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Mild-First符合要求的数据.csv",row.names = FALSE)
## "Severe-All other"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Severe-All other",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Severe-All other")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Severe-All other的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Severe-All other符合要求的数据.csv",row.names = FALSE)
## "Severe-First"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Severe-First",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Severe-First")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Severe-First的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Severe-First符合要求的数据.csv",row.names = FALSE)
## "Severe-Last"数据表中细胞自建的相关性
## 然后把各大组的pairwusecorrelation(r>=0.8,p<=0.05)分别总结出来。
## 准备数据
table(usedata2$outSC)
##
## Fatal-All other Fatal-First Fatal-Last Healthy-First
## 640 480 480 400
## Mild-All other Mild-First Severe-All other Severe-First
## 480 560 1600 880
## Severe-Last
## 880
corrdata <- usedata2[usedata2$outSC == "Severe-Last",c(1,2,3,12)]
corrdata2 <- corrdata[c("factor","arraydata")]
corrdata2$factor <- factor(corrdata2$factor)
## 计算需要的数据
healF <- use5data(data = corrdata2,dataname = "Severe-Last")
## 保存所需要的数据
## 相关系数数据
write.csv(healF$reldata,"Severe-Last的结果.csv",row.names = FALSE)
## 符合要求的相关系数数据
write.csv(healF$usereldata,"Severe-Last符合要求的数据.csv",row.names = FALSE)