Platon is Time-Intensive about platon

[oschwengers]

Hi! Actually, Platon was not designed for the analysis of metagenomes but draft genomes of bacterial isolates.
However, the workflow should in principle work fine with metagenomes, too though the initial gene prediction by Prodigal runs in a single genome instead of metagenome mode. I could add a --meta flag in an upcoming release to address this.
But the characterization workflow is indeed rather computational intensive and thus Platon will take its time.

I guess there is no simple solution to speed things up. Maybe you can just split the contig set into chunks and analyze them in parallel on a compute cluster if you have access to one?

[oschwengers]

@joshuakirsch I've added a --meta flag triggering the internal gene prediction being executed in metagenome mode (via Prodigal).
As by using --meta, there is no actual inherent relation between them anymore, which facilitates the supposed idea of separating and analyzing input contigs in smaller chunks. This will allow horizontal scalability, as well.

[joshuakirsch]

Thanks! I appreciate the quick response and fix.

[oschwengers]

You're welcome and thanks for bringing this up. This will be public in an upcoming release that will also include a db update. Until then, you can install this fix via python3 -m pip install <path-to-repo> with <path-to-repo> pointing to local copy of this repo. I'll close this for now, but please, do not hesitate to reach out and/or re-open this in any case.