A recurrence model for laryngeal cancer based on SVM and gene function clustering

Area of Science:

• Oncology
• Molecular Biology
• Bioinformatics

Background:

• Laryngeal cancer (LC) recurrence poses a significant clinical challenge.
• Identifying reliable biomarkers for LC recurrence is crucial for patient management.

Purpose of the Study:

• To analyze gene expression data to identify critical genes associated with laryngeal cancer recurrence.
• To develop a prognostic model for predicting LC recurrence.

Main Methods:

• Utilized two gene expression datasets from the Gene Expression Omnibus (GEO).
• Employed dataset GSE27020 as a training set (75 non-recurred, 34 recurred LC cases).
• Performed differential gene expression analysis, gene pair correlation analysis, and functional enrichment analysis.

Main Results:

• Identified 725 differentially expressed genes (DEGs) and analyzed gene pair correlations in non-recurred, recurred, and both LC groups.
• Revealed seven overlapping biological functions including focal adhesion and ECM-receptor interaction.
• Identified eight feature genes (PDIA3, MYH11, PDK1, SDC3, RPE65, LAMC3, BTK, UPK1B) with validated prognostic effects.

Conclusions:

• A prognostic model for laryngeal cancer was developed.
• Several critical genes were identified as potential biomarkers for LC recurrence prediction.
• These findings may contribute to improved clinical strategies for managing laryngeal cancer.