Discovering the molecular differences between right- and left-sided colon cancer using machine learning methods.

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Left-sided colon cancer (LCC) and right-sided colon cancer (RCC) exhibit clinicopathological variations.
• Key molecular differences between LCC and RCC remain unclear, with conflicting research findings.

Purpose of the Study:

• To identify distinct molecular features differentiating LCC and RCC.
• To develop accurate classification models for LCC and RCC using machine learning.

Main Methods:

• Utilized The Cancer Genome Atlas (TCGA) RNA sequencing and mutation data from 323 LCC and 283 RCC patients.
• Applied machine learning (ML) for feature selection and classification model development.
• Conducted correlation analysis between differentially expressed genes (DEGs) and mutations using logistic regression (LR).

Main Results:

• Identified 30 key mutations and 17 key gene expression features differentiating LCC and RCC using ML.
• Achieved high classification accuracy (AUC 0.8 for mutations, 0.96 for gene expression).
• PRAC1 expression and BRAF V600E mutation (rs113488022) were identified as critical features; rs113488022 significantly correlated with four genes.

Conclusions:

• Machine learning effectively identified key molecular differences between LCC and RCC.
• These molecular distinctions can accurately classify LCC and RCC patients.
• The identified differences may explain clinical variations and guide personalized treatment strategies.