Prognostic model development for classification of colorectal adenocarcinoma by using machine learning model based on...

Area of Science:

• Oncology
• Genomics
• Bioinformatics

Background:

• Colorectal cancer (CRC) is a leading cause of cancer-related mortality globally.
• Understanding the molecular mechanisms driving CRC progression is crucial for improving patient outcomes.

Purpose of the Study:

• To identify significant differentially expressed genes (DEGs) in colorectal cancer.
• To develop a machine learning (ML)-based prognostic classification model for CRC.
• To investigate the correlation between key genes and tumor immunocyte infiltration.

Main Methods:

• Analysis of CRC mRNA gene expression datasets from TCGA and GEO.
• Identification of DEGs using DESeq2 and topconfects R package.
• Feature selection using Boruta and ML model development (Random Forest).
• Survival and correlation analyses with immunocyte infiltration.

Main Results:

• 170 significant DEGs were identified from 770 CRC samples.
• A 33-feature Random Forest model achieved 100% accuracy for prognostic classification.
• GLP2R and VSTM2A were identified as significantly downregulated in tumor samples.
• These genes showed a strong correlation with immunocyte infiltration.

Conclusions:

• GLP2R and VSTM2A are significantly downregulated in colorectal cancer tissues.
• These genes are associated with tumor immunocyte infiltration and may suppress immune response.
• GLP2R and VSTM2A show potential as prognostic biomarkers for CRC.