Add Bakta as alternative to Prokka

[jfy133]

Description of feature

Also recommended by Prokka developer! nf-core module already exists

[jfy133]

As originally suggested here: https://github.com/nf-core/mag/issues/320#issuecomment-1283792256

[d4straub]

Not sure if its actually a good idea to replace Prokka, because:

Prokka is a software tool to annotate bacterial, archaeal and viral genomes

whereas Bakta is only for bacterial genomes?! So Prokka would make some more genome annotations accessible (even if inferior) than Bakta. However, probably make Bakta default and Prokka optional?

[jfy133]

Updated the header, I'm adding this already to funcscan as an alternative to Prokka/Prodigal already so should be relatively straightforward to add in as an 'option'

[jfy133]

I agree @d4straub ! Tagged it as low-priority accordingly

[plaquette]

is there an implementation of this yet - and if yes, how can it be used?

[d4straub]

There is a nf-core module for bakta in https://nf-co.re/modules/bakta, so adding it to the pipeline should be relatively easy with nf-core tools using command nf-core modules install bakta following https://nf-co.re/tools/#install-modules-in-a-pipeline

[jfy133]

It's on my TODO As well, but I'm happy for someone else to implement if they are interested.

The original Bakta database was extremely large and made it difficult to work with.

Fortunately we've got some more experience in nf-core/funcscan (where it is used), and the Bakta team have made a much smaller database which will make it easier to integrate and test :)

[prototaxites]

Now that domain classification of bins is in, it should be possible to direct only bacterial bins into Bakta by filtering on meta.domain - the only change needed would be to remove --join_prokaryotes from ext.args for TIARA_CLASSIFY, in order to enable classification to the bacterial/archaeal level rather than merging them to prokarya.

[jfy133]

Module exists