Hi,

Would you please include grch38 and other genomes into copynumber package or make it species agnostic by supplying user defined chromosome bands.

Currently pcf function only supports following genome builds.

if (!assembly %in% c("hg19", "hg18", "hg17", "hg16", "mm7", 

        "mm8", "mm9")) {

        stop("assembly must be one of hg19, hg18, hg17 or hg16", 

            call. = FALSE)

    }

One the same line: 

# It seems numericChrom function  called in pcf.r ( copynumber package ) has human specific hardcoded values for X and Y which means it only worked for human since mouse X -> 20 and Y -> 21 

numericChrom < - function (chrom) 
{
    if (!is.numeric(chrom)) {
        if (is.factor(chrom)) {
            chrom <- as.character(chrom)
        }
        chrx <- c(which(chrom == "x"), which(chrom == "X"))
        chrom[chrx] <- 23
        chry <- c(which(chrom == "y"), which(chrom == "Y"))
        chrom[chry] <- 24
        chrom <- as.numeric(chrom)
    }
    return(chrom)
}

Thanks,

Shriram

------------------------------------------------------

Senior Bioinformatician

Cancer, Ageing and Somatic Mutation Programme

The Wellcome Trust Sanger Institute

Wellcome Trust Genome Campus

Hinxton, Cambridgeshire

CB10 1SA

Email: sb43 AT sanger.ac.uk

-----------------------------------------------------

I have made changes to pcf.r function to make it species/build agnostic.

https://github.com/sb43/copynumber/tree/feature/species_agnostic

pcf function now accepts user supplied cytoband file.
All the chromosomes are now referred by its index values.
Original chromosomes names were reassigned to index before returning the processed results.
Same logic can be applied to other functions wherever required.
Let me know if this makes sense.