The predicted interactome of the human mitochondrial proteome

Despite the fundamental importance of mitochondria in cellular metabolism, the molecular function(s) of many mitochondrial proteins remain unknown. Since protein function can be inferred from their interacting partners, we repurpose the protein structure prediction algorithm AlphaFold Multimer (AFM) as a classification model to predict protein-protein interactions of the entire human mitochondrial proteome. By screening 630,003 protein pairs, we create a compendium of 2,895 previously known and newly observed interactions, which include the interacting partner(s) of 85 uncharacterized mitochondrial proteins, thereby linking them to a known biochemical pathway. Extending the AFM-based analysis to 11 diverse eukaryotes identifies evolutionarily conserved interactions among human hits, including regulators of core bioenergetic pathways. Our experiments, guided by these predictions, nominate protein interactions that form the coenzyme Q metabolon and define the mitochondrial copper delivery pathway to cytochrome c oxidase. Our compendium represents a powerful resource for the systematic, structure-based functionalization of the human mitochondrial proteome.