[Bioc-devel] Help with EdgeR running error

[Chia Jing Yi]

Hi,

Attached is the input file that I used to run EdgeR for my data set.

Ideally the input file should be able to run through the following R-script:
*library(cluster)*
*library(gplots)*
*library(Biobase)*
*library(ctc)*
*library(ape)*
*data = read.table("diffExpr.P0.0011_C1.matrix", header=T, com='', sep="\t")
*
*rownames(data) = data[,1] # set rownames to gene identifiers*
*data = data[,2:length(data[1,])] # remove the gene column since its now
the rowname value*
*data = as.matrix(data) # convert to matrix*
*cr = cor(data, method='spearman')*
*data = log2(data+1)*
*centered_data = t(scale(t(data), scale=F)) # center rows, mean substracted*
*gene_dist = dist(centered_data, method='euclidean')*
*hc_genes = hclust(gene_dist, method='complete')*
*hc_samples = hclust(as.dist(1-cr), method="complete") # cluster conditions*
*myheatcol = redgreen(75)*
*gene_partition_assignments <- cutree(as.hclust(hc_genes), k=6);*
*partition_colors = rainbow(length(unique(gene_partition_assignments)),
start=0.4, end=0.95)*
*gene_colors = partition_colors[gene_partition_assignments]*
*save(list=ls(all=TRUE), file="diffExpr.P0.1_C1.matrix.R.all.RData")*
*pdf("diffExpr.P0.001_C1.matrix.heatmap.pdf")*
*heatmap.2(centered_data, dendrogram='both', Rowv=as.dendrogram(hc_genes),
Colv=as.dendrogram(hc_samples), col=myheatcol, RowSideColors=gene_colors,
scale="none", density.info="none", trace="none", key=TRUE, keysize=1.2,
cexCol=1, lmat=rbind(c(5,0,4,0),c(3,1,2,0)),
lhei=c(1.5,5),lwid=c(1.5,0.2,2.5,2.5), margins=c(12,5))*
*try(heatmap.2(cr, col = cm.colors(256), scale='none', symm=TRUE, key=TRUE,
density.info='none', trace='none', symkey=FALSE,
Colv=TRUE,margins=c(10,10), cexCol=1, cexRow=1))*
*dev.off()*

However it stuck at the "*gene_dist = dist(centered_data,
method='euclidean')*" and gives the following error message:
*Loading required package: amap*
*Error in dist(centered_data, method = "euclidean") : *
*  negative length vectors are not allowed*
*Execution halted*

If I run with smaller data set.
The above R script worked fine.
I was thinking the error caused by huge input file?

Thanks for any advice to solve my doubt.

best regards
edge