Identifying sepsis susceptibility genes in post-surgical patients using an artificial intelligence approach.

Area of Science:

• Genomics
• Artificial Intelligence
• Medical Diagnostics

Background:

• Early sepsis detection is crucial for patient outcomes.
• Genome-wide association studies (GWAS) have identified sepsis genetic variants but face challenges with patient heterogeneity and analysis methods.
• This study focuses on identifying new sepsis susceptibility loci in post-surgical patients using an explainable AI (XAI) approach on GWAS data.

Purpose of the Study:

• To identify novel sepsis susceptibility loci in post-surgical patients.
• To apply an explainable artificial intelligence (XAI) methodology to GWAS data for sepsis prediction.
• To prioritize genetic variants associated with post-operative sepsis.

Main Methods:

• Performed GWAS on 750 post-operative sepsis patients and 3,500 controls.
• Applied a novel XAI-based methodology to GWAS-derived single nucleotide polymorphisms (SNPs) for sepsis prediction.
• Assessed functional and enrichment effects of top-ranked variants and associated genes.

Main Results:

• The XAI-GWAS approach effectively predicted post-surgical sepsis.
• Prioritized SNPs (e.g., rs17653532, rs1575081785, rs74707084) showed high contribution to sepsis prediction.
• Discovered sepsis risk loci with functional implications in gene expression, DNA replication, signaling, cell proliferation, and cardiac function.

Conclusions:

• The combination of GWAS and XAI identified loci associated with post-operative sepsis susceptibility.
• Key genes (PRIM2, SYNPR, RBSN) identified through pre-operative tests can improve risk stratification and early sepsis detection.
• Further validation in diverse cohorts is needed to confirm findings and improve patient outcomes.