#######################################################
#   Function for WAD (Weighted Average Difference) in R
#   Written by: Koji Kadota, Univ. Tokyo, 2007
#   Citation: Kadota K, Nakai Y, Shimizu K., A weighted average difference
#   method for detecting differentially expressed genes
#   using DNA microarrays. Algorithms Mol. Biol., 2008
#######################################################
WAD <- function(x, cl, dynamic_r, min_v){ 
    x.class1 <- x[(cl == 0)]
    x.class2 <- x[(cl == 1)]
    x_ave <- (mean(x.class1) + mean(x.class2))/2
    weight <- (x_ave - min_v)/dynamic_r
    statistic <- (mean(x.class2) - mean(x.class1))*weight
    return(statistic)
}

#############################
###  Average Difference   ###
#############################
AveDiff <- function(x, cl){
    x.class0 <- x[(cl == 0)]
    x.class1 <- x[(cl == 1)]
    statistic <- mean(x.class1) - mean(x.class0)
    return(statistic)
}

#############################
###  Average Fold Change  ###
#############################
AvelogFC <- function(x, cl){
    x.class0 <- x[(cl == 0)]
    x.class1 <- x[(cl == 1)]
    statistic <- log2(mean(x.class1)/mean(x.class0))
    return(statistic)
}


source('http://eh3.uc.edu/r/ibmtR.R')
library(RankProd) #パッケージの読み込み
library(samr)
library(st) 
library(limma)
library(ROC)

###############################
###############################
###  MAS-preprocessed data  ###
###############################
###############################
data <- read.table("http://affycomp.biostat.jhsph.edu/AFFY2/rafa@jhu.edu/030424.1033/hgu133.csv", header=TRUE, row.names=1, sep=",")
data[data < 1] <- 1
data <- log(data, 2)
data.cl.org <- c(0,1,0,1,0,1)
data.cl <- data.cl.org

###################################################################
###  Assignment for 42 true differentially expressed probesets  ###
data.degenes <- as.matrix(c(rep(0, nrow(data))))
rownames(data.degenes) <- rownames(data)
degenes <- c(
"203508_at", 
"204563_at", 
"204513_s_at", 
"204205_at", 
"204959_at", 
"207655_s_at", 
"204836_at", 
"205291_at", 
"209795_at", 
"207777_s_at", 
"204912_at", 
"205569_at", 
"207160_at", 
"205692_s_at", 
"212827_at", 
"209606_at", 
"205267_at", 
"204417_at", 
"205398_s_at", 
"209734_at", 
"209354_at", 
"206060_s_at", 
"205790_at", 
"200665_s_at", 
"207641_at", 
"207540_s_at", 
"204430_s_at", 
"203471_s_at", 
"204951_at", 
"207968_s_at", 
"AFFX-r2-TagA_at", 
"AFFX-r2-TagB_at", 
"AFFX-r2-TagC_at", 
"AFFX-r2-TagD_at", 
"AFFX-r2-TagE_at", 
"AFFX-r2-TagF_at", 
"AFFX-r2-TagG_at", 
"AFFX-r2-TagH_at", 
"AFFX-DapX-3_at", 
"AFFX-LysX-3_at", 
"AFFX-PheX-3_at", 
"AFFX-ThrX-3_at"
)
data.degenes[degenes, ] <- 1
###################################################################
write.table(data.degenes, "data_DEG.txt", sep = "\t", append=F, quote=F, row.names=F)


#################################
###  AD (Average Difference)  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AveDiff, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j,"\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_AD <- apply(quantile_tmp, 1, mean)
AUC_AD <- mean(AUC_tmp)

#############
###  WAD  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- apply(data[,analysis_vector == 1], 1, WAD, data.cl, dynamic_range, min(ave_vector))

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_WAD <- apply(quantile_tmp, 1, mean)
AUC_WAD <- mean(AUC_tmp)

#############################
###  WAD's "w" term only  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- (ave_vector - min(ave_vector))/dynamic_range

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_w.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_w.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_w <- apply(quantile_tmp, 1, mean)
AUC_w <- mean(AUC_tmp)

############
###  FC  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.degenes.vector <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data <- 2^data
        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AvelogFC, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        data <- log(data, 2)
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_FC <- apply(quantile_tmp, 1, mean)
AUC_FC <- mean(AUC_tmp)

############
###  RP  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- RP(data[,analysis_vector == 1], data.cl, num.perm=1, logged=TRUE, na.rm = FALSE, plot = FALSE, rand = 123)

        rank_method <- rank(apply(tmpall$RPs, 1, min), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_RP <- apply(quantile_tmp, 1, mean)
AUC_RP <- mean(AUC_tmp)

##############
###  ibmT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        design <- model.matrix(~data.cl) 
        data <- as.matrix(data)
        fit <- lmFit(data[,analysis_vector == 1], design)
        fit$Amean<-rowMeans(data[,analysis_vector == 1])
        fit <- IBMT(fit,2)
        tmpall <- fit$IBMT.t

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_ibmT <- apply(quantile_tmp, 1, mean)
AUC_ibmT <- mean(AUC_tmp)

##############
###  modT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- modt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_modT <- apply(quantile_tmp, 1, mean)
AUC_modT <- mean(AUC_tmp)

##############
###  samT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- sam.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_samT <- apply(quantile_tmp, 1, mean)
AUC_samT <- mean(AUC_tmp)

#################
###  shrinkT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- shrinkt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_MAS_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_MAS_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_shrinkT <- apply(quantile_tmp, 1, mean)
AUC_shrinkT <- mean(AUC_tmp)

tmp <- rbind(quantile_w, quantile_AD, quantile_WAD, quantile_FC, quantile_RP, quantile_modT, quantile_samT, quantile_shrinkT, quantile_ibmT)
write.table(tmp, "result_Dataset2_quantile_MAS.txt", sep = "\t", append=F, quote=F, row.names=F)

tmp <- AUC_w
tmp <- rbind(tmp, AUC_AD)
tmp <- rbind(tmp, AUC_WAD)
tmp <- rbind(tmp, AUC_FC)
tmp <- rbind(tmp, AUC_RP)
tmp <- rbind(tmp, AUC_modT)
tmp <- rbind(tmp, AUC_samT)
tmp <- rbind(tmp, AUC_shrinkT)
tmp <- rbind(tmp, AUC_ibmT)
write.table(tmp, "result_Dataset2_AUC_MAS.txt", sep = "\t", append=F, quote=F, row.names=F)  




###############################
###############################
###  RMA-preprocessed data  ###
###############################
###############################
data <- read.table("http://affycomp.biostat.jhsph.edu/AFFY2/rafa@jhu.edu/030429.1332/hgu133.csv", header=TRUE, row.names=1, sep=",")
data[data < 1] <- 1
data <- log(data, 2)
data.cl <- data.cl.org

###################################################################
###  Assignment for 42 true differentially expressed probesets  ###
data.degenes <- as.matrix(c(rep(0, nrow(data))))
rownames(data.degenes) <- rownames(data)
degenes <- c(
"203508_at", 
"204563_at", 
"204513_s_at", 
"204205_at", 
"204959_at", 
"207655_s_at", 
"204836_at", 
"205291_at", 
"209795_at", 
"207777_s_at", 
"204912_at", 
"205569_at", 
"207160_at", 
"205692_s_at", 
"212827_at", 
"209606_at", 
"205267_at", 
"204417_at", 
"205398_s_at", 
"209734_at", 
"209354_at", 
"206060_s_at", 
"205790_at", 
"200665_s_at", 
"207641_at", 
"207540_s_at", 
"204430_s_at", 
"203471_s_at", 
"204951_at", 
"207968_s_at", 
"AFFX-r2-TagA_at", 
"AFFX-r2-TagB_at", 
"AFFX-r2-TagC_at", 
"AFFX-r2-TagD_at", 
"AFFX-r2-TagE_at", 
"AFFX-r2-TagF_at", 
"AFFX-r2-TagG_at", 
"AFFX-r2-TagH_at", 
"AFFX-DapX-3_at", 
"AFFX-LysX-3_at", 
"AFFX-PheX-3_at", 
"AFFX-ThrX-3_at"
)
data.degenes[degenes, ] <- 1
###################################################################


#################################
###  AD (Average Difference)  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AveDiff, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j,"\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_AD <- apply(quantile_tmp, 1, mean)
AUC_AD <- mean(AUC_tmp)

#############
###  WAD  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- apply(data[,analysis_vector == 1], 1, WAD, data.cl, dynamic_range, min(ave_vector))

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_WAD <- apply(quantile_tmp, 1, mean)
AUC_WAD <- mean(AUC_tmp)

#############################
###  WAD's "w" term only  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- (ave_vector - min(ave_vector))/dynamic_range

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_w.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_w.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_w <- apply(quantile_tmp, 1, mean)
AUC_w <- mean(AUC_tmp)

############
###  FC  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.degenes.vector <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data <- 2^data
        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AvelogFC, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        data <- log(data, 2)
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_FC <- apply(quantile_tmp, 1, mean)
AUC_FC <- mean(AUC_tmp)

############
###  RP  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- RP(data[,analysis_vector == 1], data.cl, num.perm=1, logged=TRUE, na.rm = FALSE, plot = FALSE, rand = 123)

        rank_method <- rank(apply(tmpall$RPs, 1, min), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_RP <- apply(quantile_tmp, 1, mean)
AUC_RP <- mean(AUC_tmp)

##############
###  ibmT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        design <- model.matrix(~data.cl) 
        data <- as.matrix(data)
        fit <- lmFit(data[,analysis_vector == 1], design)
        fit$Amean<-rowMeans(data[,analysis_vector == 1])
        fit <- IBMT(fit,2)
        tmpall <- fit$IBMT.t

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_ibmT <- apply(quantile_tmp, 1, mean)
AUC_ibmT <- mean(AUC_tmp)

##############
###  modT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- modt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_modT <- apply(quantile_tmp, 1, mean)
AUC_modT <- mean(AUC_tmp)

##############
###  samT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- sam.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_samT <- apply(quantile_tmp, 1, mean)
AUC_samT <- mean(AUC_tmp)

#################
###  shrinkT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- shrinkt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_RMA_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_RMA_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_shrinkT <- apply(quantile_tmp, 1, mean)
AUC_shrinkT <- mean(AUC_tmp)

tmp <- rbind(quantile_w, quantile_AD, quantile_WAD, quantile_FC, quantile_RP, quantile_modT, quantile_samT, quantile_shrinkT, quantile_ibmT)
write.table(tmp, "result_Dataset2_quantile_RMA.txt", sep = "\t", append=F, quote=F, row.names=F)

tmp <- AUC_w
tmp <- rbind(tmp, AUC_AD)
tmp <- rbind(tmp, AUC_WAD)
tmp <- rbind(tmp, AUC_FC)
tmp <- rbind(tmp, AUC_RP)
tmp <- rbind(tmp, AUC_modT)
tmp <- rbind(tmp, AUC_samT)
tmp <- rbind(tmp, AUC_shrinkT)
tmp <- rbind(tmp, AUC_ibmT)
write.table(tmp, "result_Dataset2_AUC_RMA.txt", sep = "\t", append=F, quote=F, row.names=F)  





###############################
###############################
###  DFW-preprocessed data  ###
###############################
###############################
data <- read.table("http://affycomp.biostat.jhsph.edu/AFFY2/zhongxue@mail.smu.edu/060612.1450/hgu133.csv", header=TRUE, row.names=1, sep=",")
data[data < 1] <- 1
data <- log(data, 2)
data.cl <- data.cl.org

###################################################################
###  Assignment for 42 true differentially expressed probesets  ###
data.degenes <- as.matrix(c(rep(0, nrow(data))))
rownames(data.degenes) <- rownames(data)
degenes <- c(
"203508_at", 
"204563_at", 
"204513_s_at", 
"204205_at", 
"204959_at", 
"207655_s_at", 
"204836_at", 
"205291_at", 
"209795_at", 
"207777_s_at", 
"204912_at", 
"205569_at", 
"207160_at", 
"205692_s_at", 
"212827_at", 
"209606_at", 
"205267_at", 
"204417_at", 
"205398_s_at", 
"209734_at", 
"209354_at", 
"206060_s_at", 
"205790_at", 
"200665_s_at", 
"207641_at", 
"207540_s_at", 
"204430_s_at", 
"203471_s_at", 
"204951_at", 
"207968_s_at", 
"AFFX-r2-TagA_at", 
"AFFX-r2-TagB_at", 
"AFFX-r2-TagC_at", 
"AFFX-r2-TagD_at", 
"AFFX-r2-TagE_at", 
"AFFX-r2-TagF_at", 
"AFFX-r2-TagG_at", 
"AFFX-r2-TagH_at", 
"AFFX-DapX-3_at", 
"AFFX-LysX-3_at", 
"AFFX-PheX-3_at", 
"AFFX-ThrX-3_at"
)
data.degenes[degenes, ] <- 1
###################################################################


#################################
###  AD (Average Difference)  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AveDiff, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j,"\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_AD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_AD <- apply(quantile_tmp, 1, mean)
AUC_AD <- mean(AUC_tmp)

#############
###  WAD  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- apply(data[,analysis_vector == 1], 1, WAD, data.cl, dynamic_range, min(ave_vector))

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_WAD.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_WAD <- apply(quantile_tmp, 1, mean)
AUC_WAD <- mean(AUC_tmp)

#############################
###  WAD's "w" term only  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.sub <- NULL
tmp.class1 <- NULL
tmp.class2 <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        data.sub <- data[,analysis_vector == 1]
        tmp.class1 <- apply(data.sub[,data.cl == 0], 1, mean)
        tmp.class2 <- apply(data.sub[,data.cl == 1], 1, mean)
        ave_vector <- (tmp.class1 + tmp.class2)/2
        dynamic_range <- max(ave_vector) - min(ave_vector)
        tmpall <- (ave_vector - min(ave_vector))/dynamic_range

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_w.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_w.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_w <- apply(quantile_tmp, 1, mean)
AUC_w <- mean(AUC_tmp)

############
###  FC  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
data.degenes.vector <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data <- 2^data
        data.cl <- data.cl.org
        tmpall <- apply(data[,analysis_vector == 1], 1, AvelogFC, data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        data <- log(data, 2)
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_FC.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_FC <- apply(quantile_tmp, 1, mean)
AUC_FC <- mean(AUC_tmp)

############
###  RP  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        tmpall <- RP(data[,analysis_vector == 1], data.cl, num.perm=1, logged=TRUE, na.rm = FALSE, plot = FALSE, rand = 123)

        rank_method <- rank(apply(tmpall$RPs, 1, min), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_RP.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_RP <- apply(quantile_tmp, 1, mean)
AUC_RP <- mean(AUC_tmp)

##############
###  ibmT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org
        design <- model.matrix(~data.cl) 
        data <- as.matrix(data)
        fit <- lmFit(data[,analysis_vector == 1], design)
        fit$Amean<-rowMeans(data[,analysis_vector == 1])
        fit <- IBMT(fit,2)
        tmpall <- fit$IBMT.t

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_ibmT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_ibmT <- apply(quantile_tmp, 1, mean)
AUC_ibmT <- mean(AUC_tmp)

##############
###  modT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- modt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_modT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_modT <- apply(quantile_tmp, 1, mean)
AUC_modT <- mean(AUC_tmp)

##############
###  samT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- sam.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_samT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_samT <- apply(quantile_tmp, 1, mean)
AUC_samT <- mean(AUC_tmp)

#################
###  shrinkT  ###
quantile_tmp <- NULL
AUC_tmp <- NULL
rank_method <- NULL
analysis_vector <- c(rep(0, ncol(data)))
for(i in 1:(ncol(data)/3-1)){
        for(j in (i+1):(ncol(data)/3)){
        analysis_vector <- c(rep(0, ncol(data)))
        analysis_vector[c(i,i+ncol(data)/3, i+ncol(data)/3*2)] <- 1
        analysis_vector[c(j,j+ncol(data)/3, j+ncol(data)/3*2)] <- 1

        data.cl <- data.cl.org + 1
        tmpall <- shrinkt.stat(t(data[,analysis_vector == 1]), data.cl)

        rank_method <- rank(-abs(tmpall), ties.method = "min")
        quantile_tmp <- cbind(quantile_tmp, quantile(rank_method[data.degenes == 1]))
        AUC_tmp <- rbind(AUC_tmp, AUC(rocdemo.sca(truth = as.vector(data.degenes), data = as.vector(-rank_method), rule = dxrule.sca)))
        #cat(i, j, "\n")
    }
}
write.table(quantile_tmp, "result_Dataset2_quantile_DFW_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
write.table(AUC_tmp, "result_Dataset2_AUC_DFW_shrinkT.txt", sep = "\t", append=F, quote=F, row.names=F)
quantile_shrinkT <- apply(quantile_tmp, 1, mean)
AUC_shrinkT <- mean(AUC_tmp)

tmp <- rbind(quantile_w, quantile_AD, quantile_WAD, quantile_FC, quantile_RP, quantile_modT, quantile_samT, quantile_shrinkT, quantile_ibmT)
write.table(tmp, "result_Dataset2_quantile_DFW.txt", sep = "\t", append=F, quote=F, row.names=F)

tmp <- AUC_w
tmp <- rbind(tmp, AUC_AD)
tmp <- rbind(tmp, AUC_WAD)
tmp <- rbind(tmp, AUC_FC)
tmp <- rbind(tmp, AUC_RP)
tmp <- rbind(tmp, AUC_modT)
tmp <- rbind(tmp, AUC_samT)
tmp <- rbind(tmp, AUC_shrinkT)
tmp <- rbind(tmp, AUC_ibmT)
write.table(tmp, "result_Dataset2_AUC_DFW.txt", sep = "\t", append=F, quote=F, row.names=F)