Include MLR estimates for Pango lineages

[trvrb]

Description of proposed changes

This PR adds support for running MLR estimates on Pango lineages. Broadly it:

• Adds a variant_classifications option that currently includes nextstrain_clades and pango_lineages
• Cleans up naming to refer to sequence counts, including prepared_seq_counts.tsv
• Drops support for renaming variants via clade_to_variant.tsv (this was unnecessary for Nextstrain clades and not possible to create the same mapping file for Pango lineages)
• Adds support for models in the primary config.yaml file
• Specifies separate pivot values for clades, ie 22B (Omicron), and lineages, ie BA.5
• Updates readme with new result file locations
• Updates tests to accommodate changes to prepare-data.py.

Testing

I've tested locally, but not yet via GitHub Actions.

• [x] Checks pass

[trvrb]

Thanks for reviewing @joverlee521 and thanks for the additional cleanup. I reran GitHub Action and with it, discovered that all the rules in upload.smk needed to be updated to include {variant_classification} wildcard. I did this in the above commit and reran action. This seems to have worked as expected given that Pango lineages results are now up on S3.

I think this should be good to merge. I'll plan to do so now.