torchcell.knowledge_graphs.subset.load_gene_sets

load_gene_sets(path: str) → list[list[str]]

Load gene sets from a text file: one set per line, whitespace-separated genes.

Blank lines and # comments are ignored, so a generating script can stamp provenance into the file it emits. A relative path resolves against the installed torchcell package, because the KG build runs in a container that pip-installs torchcell from git – the repo checkout is absent there, but the package is not.