[BioC] genoset:::initGenoSet

Schalkwyk, Leonard leonard.schalkwyk at kcl.ac.uk
Wed Feb 20 20:39:23 CET 2013


Thanks, I should have checked the devel version myself.

leo


On 20 Feb 2013, at 17:27, Pan Du wrote:

> Hi Leonard
> 
> Pete has fixed this problem in the developing version of genoset.
> Thanks for reporting this!
> 
> Pan
> 
> On Wed, Feb 20, 2013 at 4:27 AM, Schalkwyk, Leonard
> <leonard.schalkwyk at kcl.ac.uk> wrote:
>> 
>> 
>> When generating a MethyGenoSet to try out the methyAnalysis package, I got the following:
>> 
>>> c <- MethyLumiM2GenoSet(b)
>>        Error in genoset:::initGenoSet(type = "MethyGenoSet", locData = locData,  :
>>          Mismatch between sampleNames and rownames of pData
>>> 
>>> all.equal(sampleNames(b), rownames(pData(b)))
>>        [1] TRUE
>> 
>> 
>> it turns out to be because of the following lines in genoset:::initGenoSet
>> 
>>        rownames(pData) = make.names(rownames(pData), unique = TRUE)
>>        if (!setequal(rownames(pData), sampleNames(ad))) {
>>            stop("Mismatch between sampleNames and rownames of pData")
>>        }
>> 
>> 
>> which guarantees that a confusing 'Mismatch' will be reported if the sampleNames start with
>> [0-9].  In the case of Illumina DNA methylation arrays this is true  of the default names.
>> 
>> Can I suggest that make.names() is run on both sampleNames and rownames (or,
>> preferably, on neither) before this test is run?
>> 
>> Leo Schalkwyk
> 



More information about the Bioconductor mailing list