Machine Learning-Based Integrative Analysis Identifies SUMOylation-Related Genes Underlying the Immune Heterogeneity...

Area of Science:

• Molecular Biology
• Bioinformatics
• Immunology

Background:

• Sepsis heterogeneity complicates diagnosis and treatment.
• The role of SUMOylation (a post-translational modification) in sepsis is understudied.

Purpose of the Study:

• To identify key genes (hub genes) associated with SUMOylation in sepsis.
• To stratify sepsis patients into distinct subtypes based on gene expression.
• To explore potential therapeutic targets for sepsis.

Main Methods:

• Integrated three GEO datasets to create a large sepsis cohort.
• Applied machine learning algorithms to screen SUMOylation-associated differentially expressed genes (DEGs).
• Utilized unsupervised consensus clustering, ssGSEA, and GSVA to analyze sepsis subtypes and their immune/functional features.
• Validated hub gene expression in a murine sepsis model (cecal ligation and puncture).

Main Results:

• Identified 43 SUMOylation-associated DEGs, pinpointing eight hub genes (TOP2B, HDAC4, NUP43, HNRNPK, BCL11A, RPA1, RORA, XRCC4) with high diagnostic potential.
• Stratified sepsis patients into two subtypes: Subtype A (immunosuppressive, high regulatory T cell infiltration) and Subtype B (hyper-inflammatory, high effector lymphocyte infiltration).
• Vorinostat identified as a potential therapeutic compound.
• Experimental validation confirmed hub gene dysregulation.

Conclusions:

• Discovered a novel eight-gene signature associated with SUMOylation, offering new diagnostic strategies and revealing sepsis heterogeneity.
• Identified two distinct sepsis subtypes with differing immunological and functional profiles.
• Emphasized SUMOylation's role in sepsis pathophysiology, paving the way for precision diagnostics and personalized therapy.