Abstract
Objective: To identify plasma protein targets with potential therapeutic implications for sepsis using genetic and proteomic data integration.
Methods: We explored plasma proteomics data from deCODE Genetics, which measured 4,719 proteins in 35,559 Icelandic individuals, and genomics data on sepsis from 361,141 participants in the UK Biobank. Utilizing cis-pQTLs as instrumental variables, we conducted Mendelian Randomization to identify circulating plasma proteins causally linked to the risk of sepsis. After adjusting for false discovery rate (FDR), the associated proteins were further analyzed through Protein-Protein Interaction analysis and Bayesian colocalization. Ultimately, protein exhibiting the strongest colocalization evidence was subjected to molecular docking to identify targeted therapeutics for sepsis.
Results: From 229 initial proteins, 27 significant proteins pass FDR correction. Among these, 11 proteins showed positive associations and 16 demonstrated negative associations with sepsis risk. Protein-Protein Interactions analysis indicated strong interactions among 15 proteins related to immune and inflammatory responses. Bayesian colocalization analysis identified GFER protein as having the strongest evidence of colocalization. GFER protein demonstrated stable binding with nicotinamide, positioning them as high-potential drug targets.
Conclusion: Our results highlight the effectiveness of integrating genetic and proteomic data to identify new therapeutic targets for sepsis. GFER protein is particularly promising candidates for further therapeutic development.