Open schifferl opened 1 year ago
Upon further investigation, the issue arises when number_reads
is curated incorrectly. In the case of the MetaCardis_2020_a
study, values in the number_reads
column all end in .0
which is read into R as NA
rather than an integer. Fixing this issue in curatedMetagenomicDataCuration will resolve the issue in curatedMetagenomicData.
Describe the bug
When data is requested using either the
curatedMetagenomicData
orreturnSamples
function with the argumentcounts = TRUE
, the assay of the object to be returned will sometimes have zero columns.To Reproduce
This bug can be reproduced with the
MetaCardis_2020_a
study as follows.Expected behavior
The code above should return a
TreeSummarizedExperiment
with 697 rows and 1831 columns.Additional context
This bug might apply to other
relative_abundance
studies, and the number to studies impacted should be verified (by downloading them with thecounts = TRUE
argument). This may be related to changes inTreeSummarizedExperiment
orSummarizedExperiment
. This issue closes waldronlab/curatedMetagenomicData#292.