This was broken in the previous commit that shuffled things around to
conform to pathogen repo guide. Get it working again taking ingest
results as input.

Dataset maintainers and description have been updated.

Refactor the workflow to handle multiple outbreaks keyed on name,
starting with the current one keyed on 'west-africa-2014'.

At least initially, there are only a few differences between outbreak
datasets.

useful for debugging

<https://www.ncbi.nlm.nih.gov/nuccore/10313991>

So `augur align` doesn't complain that NC_002549 is also in the input
sequences.

preview: <https://next.nextstrain.org/community/victorlin/nextstrain-test@eb620d9c/ebola/all-outbreaks>