Check on wrong assay?

[p-smirnov]

PharmacoGx/R/methods-summarizeMolecularProfiles.R

Line 227 in eeee68a

if(!is.null(SummarizedExperiment::assay(res, 1))) {

Im about 60% sure this is supposed to be a 2 in the if statement, it does not really make much sense otherwise. The first assay should not be null, we just put it in there.

[p-smirnov]

@ChristopherEeles When you have a chance, can you confirm my understanding?

[ChristopherEeles]

IIRC I added that condition to replicate the output of the old function for ExpressionSets, which always produced a second assay in the summary (I think it was supposed to hold the standard error of the summarized metric). As far as I could tell it was always an NA matrix with the same dimensions and row/colnames as the first assay matrix.

If we remove that block, there will be an error in the assayNames assignment on line 236. The block serves no other function than to match the output of the original function and prevent the error I mentioned. You can see there was a note in there to generalize that block for SummarizedExperiments with more than one assay.

[ChristopherEeles]

I check if the first assay is null to deal with PharmacoSets without molecular profiles, which should not have a second assay added in the summary.