Open ctb opened 3 years ago
(this is rapidly taking on some of the "hey we need generic workflows for sourmash" needs 😆 )
but, anyway, part of the point is that we can reimplement and optimize some of these steps underneath, based on expanding capabilities of sourmash/etc.
(this is rapidly taking on some of the "hey we need generic workflows for sourmash" needs laughing ) but, anyway, part of the point is that we can reimplement and optimize some of these steps underneath, based on expanding capabilities of sourmash/etc.
so sourmash becomes a more focused "power tool", and other tools/pipelines can explore and extend before it crystalizes back into sourmash? I think this is good to avoid feature-creep in sourmash, as well as maintaining sourmash liable to its API/semantic versioning changes (because more tools will be depending on it).
(just to be clear, this is kind of what I already think about the Rust/Python parts of sourmash =])
eet eez abstractionz all ze vey down
I'm thinking of doing the command line for this snakemake app differently; right now I'm not using config files much, just the sample name. How does the below sound?
each command would be successive & snakemake-ified, of course, so you could run the last one and it would do all the previous ones.