function 'strsplit': object 'GE' not found

[yangmean]

Describe the bug
Very useful package. I used my own conda environment (R verison 4.3.1) and met this problem below:

Error in h(simpleError(msg, call)) :
error in evaluating the argument 'x' in selecting a method for function 'strsplit': object 'GE' not found
Calls: convert2countM ... .makewide -> unlist -> strsplit -> .handleSimpleError -> h
Execution halted

To Reproduce

Screenshots

**Session information
samtools 1.17
STAR 2.7.11a
R 4.3.1
pigz 2.6

Additional context
Add any other context about the problem here.

[brainfo]

same issue here with both 2.9.7c or 2.9.7e coupled with R 4.1.3 or 4.2.2, STAR 2.7.10a, samtools 1.15.1. also read the commits history, but find any clues like which commit version should work.
some of the reads in {project_name}.filtered.Aligned.GeneTagged.sorted.bam have tag GE in it, e.g.:
A00821:1413:H2F53DSX7:2:2311:13810:12132 16 chr1 14409 1 78M514N37M9S * 0 0 GCTCAGTTCTTTATTGATTGGTGTGCCGTTTTCTCTGGAAGCCTCTTAAGAACACAGTGGCGCAGGCTGGGTGGAGCCGTCCGACATCAAGTGCCCACCTTGGCTCGTGGCTTTCCCGTTGCAG FFFF:FFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF:F:FFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF:FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF NH:i:4 HI:i:1 AS:i:111 nM:i:2 BX:Z:GGATATTCCACA BC:Z:GGATATTCCACA UB:Z:CAGCGCGACGGTGTGA QB:Z:FFFFFFFFFFFF QU:Z:F,FFFFFFFFFFFFFF ES:Z:Assigned3 EN:i:1 GE:Z:ENSG00000227232.5 IS:Z:Assigned3 IN:i:1 GI:Z:ENSG00000227232.5

[yangmean]

@brainfo I checked this error in scripts. It appears to be related to the absence of the 'GE' object in the "UMIstuffFUN.R" script. I'm unsure about the meaning of the 'GE' object within the '.makewide' function in this script.

[cziegenhain]

Hi, I think this is related to using a very current R version.
Please use the included conda environment that has a validated installation.
(zUMIs.sh -c -y run.yaml)

[boegel]

@cziegenhain Do you have any idea what's actually behind this error?
Obviously GE is not being defined for some reason, but where is that supposed to happen?

I'm hoping to avoid having to use the included conda environment, because it contains copies of old software for which we have optimized installations already available, so I would like to leverage those...

Is there anything I can do to help figure out what's behind this error exactly?
I'm not an R expert, but I'm happy to try and figure out the underlying cause and come up with a fix.