Nucleotide motif-guided selection of plasma microRNA biomarkers in trauma

Trauma remains a leading cause of morbidity and mortality in part due to complex pathophysiological responses. Yet our abilities to predict these changes are limited. Plasma miRNAs have been proposed as DAMPs that drive immune response and organ injury. Here, we test a panel of plasma miRNAs-selected based on next-generation sequencing and nucleotide motifs identified via a machine-learning algorithm-for their abilities to predict subclinical pathophysiological injuries. We find marked and severity-dependent increases in the miRNA biomarkers following trauma, which are closely associated with various injury markers. AUROC indicates that these biomarkers possess strong diagnostic and predictive abilities in overall trauma severity, organ injury, coagulation, endothelial activation, and inflammation. In a combined cohort of trauma and sepsis, miR-224-5p and miR-145-5p emerge as particularly effective in differentiating the two critical illnesses. These observations offer insights into potential values of the plasma miRNAs in the prediction of critical pathophysiological injury in trauma.