Hi all, it took a little while but I ended up writing a tool (really, a thin wrapper around MMSeqs2) to do this. It can assemble and identify marker genes from Illumina reads (shotgun, RNASeq, Hi-C), PacBio/ONT reads, or just search pre-assembled contigs, and includes a tool for building marker gene databases. Basically it does a fast pre-filter of a subset of reads against the marker DB, assembles them, and does a more sensitive search to return good matches from the database. You can assert an expected taxon, too, if you think you know what the species /should/ be. Very much still a work in progress but thought I'd share in case anyone else is struggling with a similar issue! https://github.com/sanger-tol/markadoros

where is the nf-core modules already?

It's not even on pip yet 😅😅😅😅