ingest: Move Nextclade configs to separate YAML

[joverlee521]

Description of proposed changes

It is unintuitive to have to remove default configs on the first run of the ingest workflow, so move the Nextclade configs to a separate config YAML that can be optionally used in nextstrain build.

Note that once the Nextclade dataset is stabilized for a pathogen, it then makes sense to move the configs into the default config.yaml

Related issue(s)

Prompted by nextstrain/docs.nextstrain.org#179

Checklist

• Checks pass

[j23414]

Separating Nextclade configs into a distinct YAML file makes the "nextclade rules" an "opt-in" action. This makes sense because we anticipate users initiating the ingest process without an initial Nextclade dataset. I agree with you that later the user can merge it with the defaults/config.yaml if they so choose.

I was curious if you'd place the nextclade config in either "defaults" or "build-configs," and it appears you've picked the "defaults" directory.

LGTM!

[joverlee521]

I was curious if you'd place the nextclade config in either "defaults" or "build-configs," and it appears you've picked the "defaults" directory.

TBH, I waffled on this for a bit! I've kept it in the defaults directory since we are pushing to include Nextclade in the core Nextstrain ingest workflows. The idea being that you should run the ingest workflow once without it, create a Nextclade workflow, then include the nextclade config in all future ingest workflow runs. It definitely doesn't work out as nicely right now (as we've learned with dengue 🙈) but my hope is this process will be streamlined in the future.