Add `mudata-to-ui` container

[NickAkhmetov]

Remaining todos:

• update requirements/requirements-dev
• update tests to include genomic profile zarr outputs/reflect changes to column names

[NickAkhmetov]

Notes:

• I made the CBB-related logic optional in case it's not always present and avoid further complicating testing by involving the adata_to_multivec_zarr function's output, especially given the open questions RE: genome assembly; if additional flexibility/test coverage are desired, I can implement them down the line, but this approach should help get this implementation out sooner.
• I made a note to create vitessce.data_utils functions for the marker gene generation/dispersion normalization functionalities so this can be shared/deduplicated between mudata-to-ui and anndata-to-ui, and will make corresponding tickets in the appropriate task trackers.
• We will update the version of anndata used in the various containers to at least 0.8.0 (ideally 0.9.2) to ensure that they're on the same version as the Azimuth pipeline/avoid any disk format mismatches - I made a JIRA ticket for this task (CAT-601).