Question: displaying clustering results in TreeView or MapleTree
gravatar for Mark Baumeister
7.9 years ago by
Mark Baumeister40 wrote:
Hi all, I am having trouble viewing the full color spectrum (I only see "red") for data that I have clustered using the hopach method and displaying in either MapleTree or TreeView. When I use the golub or kidney data sets ("Bioconductor" examples) I have no problem seeing the different colors (green - red). The data set I am using is comprised of normalized (RMA) microarray data from 6 normal vs 30 ovarian tumor tissue samples that were orignially generated from Affymetrix HG_133A chip experiments and deposited in TCGA. After normalizing the raw CEL files I select the top 200 with the highest variance before using this data for clustering. Below is the code I am using. Could anyone tell me where I might be going wrong - i.e. why my cluster files (e.g. OvarianFzy.cdt, OvarianFzy.fct, OvarianFzy.mb; OvarianTree.atr, OvarianTree.cdt, OvarianTree.gtr) when opened in MapleTree or TreeView show only a red map as opposed to the green-red gradient ? I apologize in advance if my question is not clear, I am only a beginner at using the clustering methods. Thank you, Mark # Take the 200 genes with highest variance across the arrays in my normalized eset.r.m matrix. vars <- apply(eset.r.m, 1, var) selected <- vars > quantile(vars, (nrow(eset.r.m) - 200)/nrow(eset.r.m)) esetSub <- eset.r.m[selected, ] dim(esetSub) # create a table with only the probes selected and add coresponding gene names to column 2 of table (I do this using Excel until I can figure out how to do it in R) # this is only so that I can view the genes associated with the clustered data. write.table(selected, "C:\\temp\\JHU\\MicroArray_Analysis\\Bio_Rad\\Ovarian\\batch_9\\select ed.txt") # after adding gene symbols now read it back into R (this table has all of the selected probe IDs in column 1 and all of the corresponding gene symbols in column 2) esetSub.probes.desc <- read.table("C:\\temp\\JHU\\MicroArray_Analysis\\Bio_Rad\\Ovarian\\batc h_9\\esetSub.probes.desc.txt", sep = "\t") # now change the data type to character and mode to matrix esetSub.probes.desc <- as.character(esetSub.probes.desc) esetSub.probes.desc <- as.matrix(esetSub.probes.desc) library(hopach) # compute distance matrix gene.dist <- distancematrix(esetSub, d = "cosangle") dim(gene.dist) # run hopach to cluster genes gene.hobj <- hopach(esetSub, dmat = gene.dist) gene.hobj$clust$k # plot gene distance dplot(gene.dist, gene.hobj, ord = "final", main = "Ovarian:Gene Distance", showclusters = FALSE) # bootstrap resampling bobj <- boothopach(esetSub, gene.hobj, B = 100) bootplot(bobj, gene.hobj, ord = "bootp", main = "Ovarian cancer", showclusters = FALSE) # Clustering of arrays array.hobj <- hopach(t(esetSub), d = "euclid") array.hobj$clust$k # gene clustering and bootstrap results table makeoutput(esetSub, gene.hobj, bobj, file = "C:\\temp\\JHU\\MicroArray_Analysis\\Bio_Rad\\Ovarian\\batch_9\\Ovaria\\ovarian.out", gene.names = esetSub.probes.desc[,2]) # bootstrap fuzzy clustering boot2fuzzy(esetSub, bobj, gene.hobj, array.hobj, file = "C:\\temp\\JHU\\MicroArray_Analysis\\Bio_Rad\\Ovarian\\batch_9\\Ovaria\\OvarianFzy", gene.names = esetSub.probes.desc[,2]) # HOPACH hierarchical clustering in MapleTree hopach2tree(esetSub, file = "C:\\temp\\JHU\\MicroArray_Analysis\\Bio_Rad\\Ovarian\\batch_9\\Ovaria\\data\\OvarianTree", hopach.genes = gene.hobj, hopach.arrays = array.hobj, dist.genes = gene.dist, gene.names = esetSub.probes.desc[,2]) -- Mark Baumeister [[alternative HTML version deleted]]
ADD COMMENTlink written 7.9 years ago by Mark Baumeister40
Please log in to add an answer.


Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 383 users visited in the last hour