Integrated approaches for discovery and functional annotation of proteins of unknown function

Proteins of unknown function (PUFs) remain a persistent blind spot in molecular biology. Emerging evidence implicates many PUFs in crucial but poorly characterised roles in biomedical contexts, particularly cancer and infectious diseases. Here, we explore integrative strategies combining high-throughput experimental platforms with computational models to address this gap. We outline how functional insights can be derived across a molecular hierarchy, spanning individual proteins, interaction networks, and transient assemblies, and evaluate the distinct opportunities and challenges faced at each level. Framing these advances within a systems biology lens, we argue that characterising PUFs could redefine therapeutic discovery pipelines. We call for data-driven discovery methods and community efforts to support reproducible, scalable annotation of the 'dark' proteome.