Question on use with MAGs

[AstrobioMike]

Hi there!

Thanks for all the hard work and documentation!

I might be just confusing myself somehow or overlooking something, but it seems that here on the quick start page (https://eukulele.readthedocs.io/en/latest/quickstart.html#metagenome-assembled-genome-mag-annotation), it talks about using input nucleotide fastas for MAGs and doesn't mention that being a problem. But then on the using eukulele page here (https://eukulele.readthedocs.io/en/latest/running-eukulele.html#metagenome-assembled-genomes-mags) it says we shouldn't use nucleotide contigs, and should predict proteins first.

Am I overlooking something here, or maybe you folks learned from use/testing it's not a good idea to use on MAG contigs and just haven't been able to update the docs everywhere yet?

Thanks! -Mike

[akrinos]

Hi! Thanks so much for checking out our tool!

We indeed recommend that protein prediction be done before using EUKulele. For METs, we provide the option to use blastx on nucleotide files rather than blastp (using either BLAST or DIAMOND), but for MAGs we recommend that the protein prediction process occur outside of EUKulele, as the only protein prediction tool that we have built into the software is TransDecoder.

Of course, always open to feedback!

Arianna

[AstrobioMike]

Thanks, Arianna!

Maybe add some wording to this one, so the recommendation is consistent across?

Cheers :) -Mike