Interpretable machine learning-driven multi-omics risk stratification and drug repurposing nominates Treg/Th17 with gluconeogenesis/lactylation integration as a prognostic and druggable biomarker for glioblastoma patients - Takeaways - MDSpire

Interpretable machine learning-driven multi-omics risk stratification and drug repurposing nominates Treg/Th17 with gluconeogenesis/lactylation integration as a prognostic and druggable biomarker for glioblastoma patients

  • By

  • Siqi Xie

  • Weiming Chen

  • Bing Zhang

  • Shangeng Weng

  • July 14, 2026

Share

  • 1

    Dysregulation of Treg/Th17 balance and gluconeogenesis/lactylation contributes to glioblastoma progression.

  • 2

    The study identified Treg/Th17 and gluconeogenesis/lactylation-associated DEGs for glioblastoma patients using multi-omics analysis.

  • 3

    A TGL-associated risk stratification model was developed and validated using Lasso-cox regression analysis.

  • 4

    ODC1 was identified as an up-regulated TGL-related regulator involved in glioblastoma pathogenesis.

  • 5

    THZ-2-102–1 was proposed as a potential drug targeting ODC1 for glioblastoma treatment.

Original Source(s)

Interpretable machine learning-driven multi-omics risk stratification and drug repurposing nominates Treg/Th17 with gluconeogenesis/lactylation integration as a prognostic and druggable biomarker for glioblastoma patients

  • by Siqi Xie, Weiming Chen, Bing Zhang, Shangeng Weng

  • July 14, 2026

Related Content

A novel H2A-A127 variant is associated with human cancer and enhances tumor-related phenotypes in Drosophila melanogaster models

CD19-negative non-IgM type lymphoplasmacytic lymphoma: a case report and literature review

Machine learning-driven identification of DLL3 as a molecular target and development of a DLL3-binding cyclic peptide for glioblastoma