The diagnostic value of intestinal flora in sepsis and the protection of bacterial metabolites against intestinal damage in sepsis were analyzed based on machine learning

sepsis leads to intestinal mucosal injury, bacterial translocation, and further aggravation of intestinal and remote organ injury. And intestinal injury significantly can also contribute to critical illness, sepsis and multiorgan failure. We collected fecal samples from the sepsis and the healthy . Through intestinal flora 16sRNA sequencing , intestinal metabonomic analysis and machine learning, the result shows that Intestinal flora and metabolites of intestinal flora can not only reflect the severity of sepsis, but also have good diagnostic value for sepsis. Through transcriptome sequencing and signaling analysis of of mononuclear cells in peripheral blood (PBMC)of patients with sepsis, we found that Th17 cells had significant differences in sepsis. Through the in vivo experiments, we find that L-valine, a metabolite of bacterial flora, can protect sepsis-induced intestinal injury, inhibit inflammatory response, and inhibit the expression level of Th17 cells, which provide a new strategy for the treatment of intestinal injury induced by sepsis.