Objective:

To gain insights into the mechanisms of Treg/Th17 balance and gluconeogenesis/lactylation in glioblastoma (GBM) progression.

Approach:

• Data Integration: Integrated GBM cerebral public bulk profiles from GEO database with gluconeogenesis and lactylation gene lists to identify shared DEGs.
• Risk Stratification Model: Developed a TGL-associated risk stratification model using Lasso-cox regression analysis in TCGA-GBM training cohort and validated in GEO dataset.
• Heterogeneity Assessment: Assessed immune and intratumoral heterogeneity between high-risk and low-risk groups.
• Hub Gene Identification: Utilized SHAP for interpretability of Lasso-cox regression and identified a TGL-associated hub gene.
• Therapeutic Strategy: Performed ridge regression and molecular docking to identify optimal therapeutic strategies targeting the hub gene.

Key Findings:

• Integrated TGL can guide risk stratification and prognostic model construction for GBM patients.
• ODC1 is identified as an up-regulated TGL-related regulator involved in GBM pathogenesis.
• THZ-2-102–1 is proposed as a potential drug targeting ODC1 for GBM treatment.

Interpretation:

The study integrates TGL-associated gene signatures in GBM patient risk stratification and therapeutic frameworks using machine learning and multi-omics.

Conclusion:

This study integrates TGL-associated gene signatures in GBM patient risk stratification and therapeutic frameworks.