Use multiple consolidated databases

[boasvdp]

Hi Jim, thanks for the amazing tool. We're starting to use it more and more within our team and really like it so far.

We use GTDB as the standard database and although the database is great, it's collapsed a couple of clinically relevant species into comprehensive species clusters. We end up having to supplement GTDB to still be able to differentiate several species of interest.

The new consolidated database functionality in v0.3 is great to reduce our inode footprint, but it means we now have to rebuild the whole GTDB set of representative genomes even if we only want to add a handful of new genomes.

Would it be possible to have some sort of middle ground approach, where the a user can specify multiple consolidated databases? This would really facilitate using skani in a more modular way, e.g. swapping out databases for different analyses or easily adding new genomes without rebuilding the whole set.

I have little understanding of the internal workings of skani so I have no idea whether this is possible or how much work this would require. Feel free to close the issue if this doesn't make sense!

[bluenote-1577]

Hi Boas,

Much thanks for the kind words. I think what you suggest is very possible in practice. Essentially, something like

skani search --dbs database1 database2 ... ?

I think it would not be too hard to basically wrap this as skani search -d database1; skani search -d database2 in the intermal code and aggregate results. Let me know if that sounds like what you're requesting.

I'll add this to the TODO list. I'm a bit busy for the next month, so maybe after that I'll have time. Thanks!

[boasvdp]

That'd be perfect, thanks for putting this on the todo!