Integrative metabolite-protein interaction networks reveal potential pathways and biomarkers in sepsis.

Sepsis is a significant factor in morbidity and mortality, and the list of biomarkers that could be used to help identify and manage it is limited. We have herein, combined a set of 50 metabolites related to sepsis with a set of 171 proteins to build a protein-protein interaction network and then clustered the proteins included in the interaction network, performed pathway enrichment analysis, and provided a significance score (SScore) to rank the genes. The SScore is determined as the minus logarithm based 10 of the smallest false discovery rate-adjusted p-value in all the clusters, where a protein is located; the larger the result is the more the evidence of the relevance of the protein to sepsis. The use of this framework resulted in the prioritization of 125 higher-ranking Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that included known mechanisms, which is complement and coagulation cascades and platelet activation, and novel mechanisms, including sphingolipid metabolism and glycosylphosphatidylinositol (GPI)-anchor biosynthesis. The SScore highlighted existing sepsis-relevant genes, including, but not limited to, APOA1 and A2M, and the appearance of new candidates, including, but not limited to, IL18RAP and THBS1, capable of being used as previously unknown biomarkers. Such results provide a ranked list of pathways and genes that may guide future research on biomarker development and treatment of sepsis.