Protein science has a scale problem: databases like UniProt contain millions of protein sequences, but only a tiny fraction have reliable experimental annotations. That gap […]