[BioC] split(SE, seqnames(SE)) doesn't work anymore

Martin Morgan mtmorgan at fhcrc.org
Sat Jan 26 23:29:03 CET 2013


On 01/25/2013 12:19 PM, Tim Triche, Jr. wrote:
> Also, since I almost immediately realized I should try it,
>
> R> split(sortedBloodCells, seqnames(rowData(sortedBloodCells)))$chr9
> Error in as.vector(x, mode) : invalid 'mode' argument
>
> So... that didn't resolve it.
>
> Thanks anyone with insight into this.

Hi Tim -- I can't get a working test case in the release or devel branch, so I 
wonder if this ever worked (it would be useful to know, so as to understand what 
changed)? It's now implemented in GenomicRanges 1.11.26

   example("SummarizedExperiment")

and then

 > split(sset, seqnames(sset))$chr2
class: SummarizedExperiment
dim: 150 6
exptData(0):
assays(1): counts
rownames: NULL
rowData metadata column names(0):
colnames(6): A B ... E F
colData names(1): Treatment

Martin

>
> --t
>
>
>
>
> On Fri, Jan 25, 2013 at 12:18 PM, Tim Triche, Jr. <tim.triche at gmail.com>wrote:
>
>> Hi all,
>>
>> Were there recent changes in GenomicRanges that might cause this?
>>
>> R> class(sortedBloodCells)
>> [1] "SummarizedExperiment"
>> attr(,"package")
>> [1] "GenomicRanges"
>>
>> R> showMethods("split")
>> Function: split (package base)
>> x="ANY"
>> x="GRanges"
>>      (inherited from: x="Vector")
>> x="RangedData"
>> x="SummarizedExperiment"
>>      (inherited from: x="ANY")
>> x="Vector"
>>
>> R> split(rowData(sortedBloodCells),
>> seqnames(rowData(sortedBloodCells)))$chr9
>> GRanges with 9861 ranges and 0 metadata columns:
>> ## works fine on the rowData GRanges
>>
>> R> split(sortedBloodCells, seqnames(sortedBloodCells))$chr9
>> Error in as.vector(x, mode) : invalid 'mode' argument
>> ## used to be delegated via MM's seqinfo fixes, but now... ?
>>
>> R> packageVersion('GenomicRanges')
>> [1] '1.11.24'
>>
>>
>> Thanks,
>>
>> --t
>>
>
>
>


-- 
Computational Biology / Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109

Location: Arnold Building M1 B861
Phone: (206) 667-2793



More information about the Bioconductor mailing list