Protein-Protein Interaction Network Mapping

Abstract

Protein-protein interaction networks encode the physical and functional relationships that govern cellular signalling, metabolism, and gene regulation, and mapping these networks comprehensively is foundational for systems biology and drug discovery, yet the completeness, accuracy, and accessibility of PPI maps vary widely across experimental and computational mapping approaches. We evaluated 214 PPI network mapping programmes across centres in Italy, Sweden, and Spain between 2016 and 2021, spanning five mapping methodology categories: high-throughput yeast two-hybrid screens, affinity purification coupled with mass spectrometry, proximity labelling proteomics, computational prediction from sequence and structure, and literature-curated database aggregation. A PPI Mapping Quality Index (PMQI) was constructed from five sub-scores -- interaction detection sensitivity, false-positive control, network coverage breadth, functional validation depth, and data accessibility and standardisation -- with weights from regression against sustained adoption into active interactome research or drug discovery pipelines. PMQI correlated with adoption at r = +0.83 and discriminated adopted from non-adopted programmes with an AUC of 0.880. Affinity purification mass spectrometry scored highest (mean PMQI 0.822), while computational prediction trailed at 0.596. Only 34.6 percent of programmes exceeded the 0.75 threshold. Interaction sensitivity carried the largest regression weight (beta = +0.278), followed by false-positive control (beta = +0.230)