Hi @Daniel Lundin, I have run the

sbdi-phylomarkercheck

workflow three times: twice using Docker and once using Singularity. I get the same results for the

filtergappy/arc-ssu-r232-fwd.gapfiltered.fna

in all three runs. I ran the workflow on two independent computing systems (GCP; local HCP). Unfortunately, these deviate from the results on FigShare for GTDB R232. Specifically, I've been tracking the sequence

GB_GCA_902827205.1~LR792634.1

which is filtered out in my runs, but is present in the

sbdi-gtdb-sativa.r11rs232-1.20genomes.fna

file. Any idea what could account for this difference? Any chance sequences are added in later in the workflow as I am checking results at the filtergappy step? I'm running Nextflow v25.10.4 with the following command:

nextflow run biodiversitydata-se/sbdi-phylomarkercheck -r v1.0.2 -profile docker -resume -params-file arc-ssu-r232.params.yaml

. I have mirrored your two param files (general Nextflow params and archaeal specific params). I am using the HMM model you sent through. Hard to imagine that our input 16S sequence file differs though perhaps you can send me through the exact file you are using. Are you in a position to clear your Nextflow cache and re-run the archaeal workflow for R232 to confirm you get the same results when doing a fresh pull of all dependencies? My concern is that perhaps v1.0.2 on GitHub is not the code that was run on your system given how Nextflow can cache containers and data files.