Pathway and network analysis in glioma with the partial least squares method

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Gene expression profiling is crucial for understanding glioma biology.
• Prior analyses often overlooked complex biological and environmental factors.
• Distinguishing between high-grade glioma subtypes (III vs. IV) requires advanced analytical methods.

Purpose of the Study:

• To investigate gene expression disparities between grade III and IV gliomas.
• To identify molecular pathways and genes associated with glioma prognosis.
• To leverage advanced statistical methods for a more comprehensive analysis.

Main Methods:

• Utilized the Gene Expression Omnibus (GEO) database for glioma expression data.
• Applied partial least squares (PLS) based analysis using R statistical software.
• Conducted survival and network analyses to identify prognostic markers and key genes.

Main Results:

• Identified 1,378 differentially expressed genes between grade III and IV gliomas.
• Four prognostic pathways identified: Prion diseases, colorectal cancer, cell adhesion molecules (CAMs), and PI3K-Akt signaling.
• Two significant hub genes, ELAVL1 and FN1, were pinpointed for their association with glioma.

Conclusions:

• The study provides novel insights into glioma pathogenesis and prognosis.
• Identified molecular targets may support future therapeutic strategies for gliomas.
• PLS-based analysis offers a robust approach for complex gene expression studies in oncology.