Closed lwaldron closed 7 years ago
If the assay column is required to be a factor then this needs a validity check, and updateObject() should include this correction.
BTW, this came up with Charlotte Soneson's nice page of MultiAssayExperiments containing gene and transcript-level RangedSummarizedExperiments for matched single-cell RNA-seq: http://imlspenticton.uzh.ch:3838/conquer/
Thanks for reporting this. Fixed.
Previously, sampleMap(mae)$assay was a character but now it must be a factor, otherwise any subsetting removes all data. See the attached example.
miniconquerMAE.rda.zip