Check the batch effect by date

date2col <- function(date.list)
{
clr.template = c("red", "orange", "yellow", "green", "cyan", "blue", "purple")
num.dates <- length(date.list)
clr.list <- vector()
clr.list[1] <- "red"
c.index <- 0
for (i in 2:num.dates) {
if(date.list[i] == date.list[i-1]) {
clr.list[i] = clr.list[i-1]
} else {
c.index <- c.index+1
clr.list[i] = clr.template[c.index %% 7+1]
}
}
return(clr.list)
}



setwd("E:\\CEL\\GastricCancer\\AU\\PM_data_new\\Gastric_Affy_files\\Tumors")
data <- read.table(file="AU_GC70.rma.txt", header=T, row.names=1)
data.ctrl <- data[54614:54675, ]
library("gplots")


#data <- t(scale(t(data.ctrl), scale=T)) #standardized by row(gene)
#data[data < -3] <- -3
#data[data > 3] <- 3

data <- sweep(data.ctrl, 1, apply(data.ctrl, 1, median)) #just median centered

my.color <- c("8/4/2004","8/4/2004","11/18/2004","11/18/2004","11/25/2004","11/25/2004","11/25/2004","11/25/2004","11/25/2004","11/25/2004","11/25/2004","11/26/2004","11/26/2004","11/26/2004","11/26/2004","11/26/2004","12/2/2004","12/2/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","12/3/2004","1/14/2005","1/14/2005","1/14/2005","1/14/2005","1/14/2005","1/14/2005","2/17/2005","2/17/2005","2/17/2005","2/17/2005","2/17/2005","2/17/2005","2/17/2005","2/25/2005","2/25/2005","3/4/2005","3/4/2005","3/4/2005","3/18/2005","3/18/2005","3/23/2005","4/8/2005","4/8/2005","4/8/2005","4/8/2005","4/8/2005","4/8/2005","4/28/2005","4/28/2005","4/28/2005","4/28/2005","4/29/2005","4/29/2005","4/29/2005","4/29/2005","5/19/2005","5/24/2005","5/24/2005","6/22/2005","6/22/2005","6/22/2005","6/22/2005","6/22/2005")
#my.color <- rep("black",dim(data)[2])
my.color <- date2col(my.color)
hm<-heatmap.2(as.matrix(data), col=greenred(75), scale="none", dendrogram="none", Rowv= T, Colv=F, ColSideColors=my.color, key=TRUE, symkey=FALSE, density.info="none",trace="none", cexRow=0.75,cexCol=0.75)
pdf(file = "Batch_in_CtrlGenes.pdf", width=10, height=10)
#pdf(file = "Batch_in_CtrlGenes.pdf")
hm<-heatmap.2(as.matrix(data), col=greenred(75), scale="none", dendrogram="none", Rowv= T, Colv=F, ColSideColors=my.color, key=TRUE, symkey=FALSE, density.info="none",trace="none", cexRow=0.75,cexCol=0.75)
dev.off()