[BioC] qaGeneST bug

[Dick Beyer]

Hi all,

I am running ArrayTools qaGeneST and got an error.  Upon looking at the underlying code, it seems there is an assumption being made that doesn't need to be.

Here is the code snippet:

qaGeneST
function (object, parameters, QC, mydir = getwd(), outputFile = "QA.html")
{
     require("R2HTML")
     targetsFile <- pData(object)
     expr <- exprs(object)
     num <- match(parameters, colnames(targetsFile))
     if (length(num) < 1 | sum(is.na(num)) > 0)
         stop("Please check your parameter names, which must match the \n
                 \n  colnames of phenotype files.")

     targetsFile <- targetsFile[, num, drop = F]
     nchip <- ncol(expr)

     targetSort <- targetsFile[do.call("order", targetsFile[,
         num, drop = F]), , drop = F]

The parameters argument specifies the column names in the pData(object) pheno data.  This code will only work if the column names specified in the parameters arg are the first through n columns.

If, for example, the desired column names are from columns 2 and 3, then the line 
targetsFile <- targetsFile[, num, drop = F]
creates a 2 column matrix, but then the line that starts with targetSort is going to error out because it is looking for columns 2 and 3 in this 2 column file.

Easy enough to work around, but I thought the package maintainers might want to know.

Cheers,
Dick

*******************************************************************************
Richard P. Beyer, Ph.D.	University of Washington
Tel.:(206) 616 7378	Env. & Occ. Health Sci. , Box 354695
Fax: (206) 685 4696	4225 Roosevelt Way NE, # 100
 			Seattle, WA 98105-6099
http://depts.washington.edu/ceeh/ServiceCores/FC5/FC5.html
http://staff.washington.edu/~dbeyer