Apply Circular Binary Segmentation Using The Dnacopy Library

Question

[desmoguga]

I'm using Varscan2 to find CNV on cancer exomes. In the recommended workflow suggested by Varscan2 there is this code to apply the segmentation:

library(DNAcopy)

cn <- read.table("your.cn.file",header=F)

CNA.object <-CNA( genomdat = cn[,6], chrom = cn[,1], maploc = cn[,2], data.type = 'logratio')

CNA.smoothed <- smooth.CNA(CNA.object)

segs <- segment(CNA.smoothed, verbose=0, min.width=2)

segs2 = segs$output

write.table(segs2[,2:6], file="out.file", row.names=F, col.names=F, quote=F, sep="\t")

I think there is a mistake in the field genomedat=cn[,6]. It should be the field 7 of the copynumber file made by Varscan. Am I right? Thanks in advance

Answer 1

[dankoboldt]

To answer the original question, yes, you're right! The "genomdat" variable should be set to the field with the log2-based copy number value, so if that's in column 7, please use it.

Answer 2

[Irsan]

At first, this is more a Bioconductor related question (although you might also get solutions here). You can post your question at the Bioconductor mailing list. Second, why do you think there is a mistake, is there an error message? Third, if you ask help about R/Bioconductor errors it is helpful to post the output of

sessionInfo()

and

traceback()

Then, if you are right that the problem is in the genomedat=cn[,6] it is helpful to give more information about cn[,6] for example

str(cn[,6]) 
head(cn[,6])

cn[,6] should be a vector or matrix where each row is a genomic marker (for example "probe") and each column is a sample. Obviously the values in the matrix should be copy number estimates

At last, I would recommend to use fastseg in stead of DNAcopy because it is much faster. Especially when you have multiple samples and many genomic markers this becomes more important