recount: inconsistency between pheno table structure for gene and transcript datasets
1
0
Entering edit mode
@jacquesvan-helden-12519
Last seen 16 months ago

Dear recount team,

First of all, congratulations this great resource, which I am regularly using for training and research.

I have a problem with several studies (e.g. SRP042620), where the structure of the pheno table extracted from RData files downloaded from recount differs between transcript and gene datasets.

For the gene dataset, the characteristics field is a CompressedCharacteristics (as expected)

> phenoTable <- colData(rse_gene) ## phenotype per run
> class(phenoTable$characteristics) [1] "CompressedCharacterList" attr(,"package") [1] "IRanges"  For the transcript dataset, it is a vector > phenoTable <- colData(rse_tx) ## phenotype per run > class(phenoTable$characteristics)
[1] "character"


Is there a way to fix this ?

Many thanks,

Jacques van Helden

software error recount • 128 views
0
Entering edit mode
Last seen 12 days ago
United States

Hi,

This is a duplicate question of https://support.bioconductor.org/p/116480/ that includes the answer at https://support.bioconductor.org/p/116480/#124335

The data hasn't changed since then, so the answer should still apply (which you reported as working in https://support.bioconductor.org/p/116480/#125723)

Best, Leo