Early Identification and Prognosis Prediction of Sepsis Through Multiomics

Early Identification and Prognosis Prediction of Sepsis Through Multiomics

This study aims to integrate multi-omics data and clinical indicators to reveal pathogen-specific molecular patterns in patients with sepsis and establish prognostic prediction models through multiple machine learning algorithms.

This study aims to quantify the plasma metabolome, single nucleotide polymorphisms (SNPs) of exons and immunocytokines of septic patients with different pathogen infections and prognostic outcomes. Multi-omics data, cytokines, and clinical indicators will be integrated through multiple machine learning algorithms to reveal pathogen-specific molecular patterns and multi-dimensional prognostic prediction models.

Inclusion Criteria:

• Patients with sepsis or septic shock who meet the diagnostic criteria (2016 sepsis 3.0 standard);
• Age 18～85 years old.

Exclusion Criteria:

• ICU stay of the subjects less than 72 hours;
• Female subjects who are pregnant;
• The subjects not sure if infected;
• The subjects performed CPR;
• The subjects suffer from chronic renal disease;
• The subjects with incomplete clinical data.