Hey all! I have a question regarding host removal in Taxprofiler. Currently, I am working with metagenomic data from three animal species. These three species naturally have very large genomes. At the moment, I have to concatenate all three genomes + the human genome to be able to run all samples in one go. This naturally takes a long time, even if I do the indexing beforehand. However, would i perhaps be possible in the future to add the host removal reference in the samplesheet per sample? That way everything can be run in one go, but with different references for host removal. Not sure what has been planned here, but that was just a thought i had!

Hrmm I see your point. I think the reason why I didn't consider that in the past is I personally would not want to compare taxonomic profiles from samples of different species... (in the sense the primary output of taxprofiler is meant to be standardised outputs you can make a consensus or compre between...)

Thanks for your feedback! In this case we aren't necessarily comparing the different species directly, I was just lazy and wanted to run everything in one go, then using the results separately for each species. In this case it would be useful to specify the reference for each one!

Haha fair enough. Like theoretically we could, so feel free to make an issue. One could argue there is a good provenance case for embeddkng the path to the reference in the samplesheet, same as we want to put in the taxdump files associated with a database. Just it can be a slippery slope as if we add the reference, then we need another column for a index file, and then it's not just one index file but a directory for bowtie 2 and then another for a minimap mmi

Thanks!

So will need to consider the balance there

Sure thing, I'll add it as an issue soon, then you can consider if it is worth the work!

(is it worth the coding pain)