As datasets grow they are increasingly split across many files. We need to be able to flexibly and cleanly break up taxonomies so that we can generate taxonomies corresponding to specific matrix (h5ad) files.
As taxonomies are single inheritance hierarchies, they can be cleanly split by specifying a node. All subnodes can then be copied into a separate taxonomy file.
Use case:
As taxonomies are single inheritance hierarchies, they can be cleanly split by specifying a node. All subnodes can then be copied into a separate taxonomy file.