Dear List members I used the following command to generate heatmap of expression value of differentially expressed genes. library("RColorBrewer") d <- read.delim("DE4018.txt", header=TRUE) x <- as.matrix(d) X_scaled <- t(apply(x, 1, scale)) hr <- hclust(as.dist(1-cor(t(X_scaled), method="spearman")), method="complete") hc <- hclust(as.dist(1-cor(X_scaled, method="spearman")), method="complete") hmcol<-brewer.pal(11,"RdBu") pdf(file = "heatmap20120517_4.pdf", height = 14, width = 7) heatmap(X_scaled, Rowv=as.dendrogram(hr), Colv=as.dendrogram(hc), col=hmcol, scale="row") dev.off() The generated map can not satisfy me fully. Because with using the dendrogram, samples of the same condition do not cluster together. In other words, samples from two conditions mixed up. My intention is to get the replicates from one condition cluster together. Could anyone please help me modify this code? Best regards Li

that depends a lot from your data, you can't "force" samples to go together (or at least you shouldn't) check your original data to see whether they go together if you use apply dimensionality reduction analysis (i.e. PCA) also, try different clustering methods, they will probably all lead to different results. Seb On Fri, Nov 30, 2012 at 3:57 PM, Wang, Li <li.wang@ttu.edu> wrote: > Dear List members > > I used the following command to generate heatmap of expression value of > differentially expressed genes. > > library("RColorBrewer") > > d <- read.delim("DE4018.txt", header=TRUE) > x <- as.matrix(d) > X_scaled <- t(apply(x, 1, scale)) > hr <- hclust(as.dist(1-cor(t(X_scaled), method="spearman")), > method="complete") > hc <- hclust(as.dist(1-cor(X_scaled, method="spearman")), > method="complete") > hmcol<-brewer.pal(11,"RdBu") > pdf(file = "heatmap20120517_4.pdf", height = 14, width = 7) > heatmap(X_scaled, Rowv=as.dendrogram(hr), Colv=as.dendrogram(hc), > col=hmcol, scale="row") > dev.off() > > The generated map can not satisfy me fully. Because with using the > dendrogram, samples of the same condition do not cluster together. In other > words, samples from two conditions mixed up. My intention is to get the > replicates from one condition cluster together. Could anyone please help me > modify this code? > > Best regards > Li > _______________________________________________ > Bioconductor mailing list > Bioconductor@r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: > http://news.gmane.org/gmane.science.biology.informatics.conductor > [[alternative HTML version deleted]]