Tracking this possibility here since our deep integration with TBprofiler and use of its intermediate files is quite complicated at the moment, but sometimes users just wish to detect mixed-infections.
This publication seems promising and I'm evaluating this independently https://www.nature.com/articles/s42003-025-07705-9 to see how well it performs.
If the results are good then I intend to add it as an optional subworkflow, similar to TBPROFILER and NTMPROFILER on FASTQ files, without any pre/post-processing of data.
Happy to hear thoughts here.