Consensus Of Hierarchical Clustering

Question

[Diana]

Hi all!

I am permuting (1000 times) the columns of my matrix containing gene expression data and then using hierarchical clustering to cluster the data in R. I don't understand how to make a consensus tree out of my results and plot it. Someone told me about consensus function in maanova but it works only for a specific type of object with fitted ANOVA model so I don't want to use that. Can anyone suggest anything else in R? Here's my code:

make.permuted.clust <- function(i)
{ 
  permuted <- data.matrix[,sample(ncol(data.matrix), 12, replace=TRUE)]
  d = dist(permuted, method = "euclidean", diag = FALSE, upper = FALSE, p = 2)
  clust = hclust(d, method = "complete", members=NULL)
  clust
}
all.clust <- lapply(1:1000, make.permuted.clust)

Thanks!!

Answer 1

[Zev.Kronenberg]

You should check out the package PVCLUST. It can constuct confidence intervals around clustered groups.

PVCLUST WEB

You could also try bootstrapping your original cluster using boot.phylo in the APE package. It's designed for phylogentics but you can use any distance function to generate a tree.