Provide a way for cladetime users to specify a nextclade dataset version for custom clade assignments

[bsweger]

Background

Currently, the CladeTime.assign_clades method determines which Nextclade SARS-CoV-2 dataset version to use for clade assignment by using the tree_as_of date:

1. Get the version of https://nextstrain-data.s3.amazonaws.com/files/ncov/open/metadata_version.json that corresponds to tree_as_of
2. The nextclade_dataset_version key of the above file is the dataset version that contains the reference tree used to assign clades in the Nextclade pipeline run on tree_as_of

This process ensures reproducibility when generating target data for variant-nowcast-hub. However, it means we can only do custom clade assignments back through August, 2024, when Nextstrain began publishing metadata_version.json

Request

For people who want to do custom clade assignments for dates prior to 2024-08-01, it would be helpful if Cladetime could accept a user-supplied SARS-CoV-2 dataset version instead of a tree-as-of date. In other words, allow users to supply the dataset version instead of having Cladetime look it up.

Definition of done

• CladeTime.assign_clades works with a user-supplied SARS-CoV-2 dataset version in addition to its existing capability of retrieving the dataset version via tree_as_of