Evolutionary rate covariation across malaria parasite species enables inference of protein interactions

Despite publication of the Plasmodium falciparum reference genome over 20 years ago, one-third of its genes remain functionally unannotated, and information is limited for many others. Proteins that act in the same pathway or complex tend to experience similar shifts in evolutionary pressure, such that correlated constraints across species can indicate co-functional proteins. To investigate this connection, we calculated relative evolutionary rates across the genome on a phylogeny of 22 Plasmodium species and assigned each protein-protein pair a score representing the strength of evolutionary rate covariation (ERC). We show that known pathways and interacting proteins across lifecycle stages in Plasmodium have strong ERC signals. By scanning genome-wide for additional proteins showing high ERC with established interacting proteins, we find enrichment of stage expression and physically interacting protein pairs supporting new candidate functions for proteins. More generally, we demonstrate the utility of ERC to prioritize proteins for hypothesis-driven functional follow up by showing that a protein with little functional characterization (PF3D7_0811600) shows high ERC and co-localizes with high molecular weight rhoptry proteins 2 and 3 (RhopH2 and RhopH3), which form an ion channel that enhances parasite permeability of infected red blood cells. The ERC matrix can be queried to extract Plasmodium proteins showing high ERC with any protein of interest to prioritize candidate genes and accelerate discovery of novel functional connections.