Identification of Key Osteoarthritis-Associated Genes Based on DNA Methylation

Area of Science:

• Epigenetics
• Computational Biology
• Genomics

Background:

• Osteoarthritis (OA) diagnosis and molecular understanding are limited.
• DNA methylation plays a crucial role in OA pathogenesis.
• Novel biomarkers are needed for early detection and biological interpretation.

Purpose of the Study:

• To develop an integrative computational framework for identifying OA-associated genes and DNA methylation biomarkers.
• To validate these biomarkers for diagnostic accuracy and biological insights.
• To explore novel epigenetic mechanisms in OA.

Main Methods:

• Integrative computational framework combining statistical analysis, machine learning, and deep learning.
• Differential methylation analysis and selection of CpG sites near OA-related genes.
• Support Vector Machine Recursive Feature Elimination and Random Forest for feature refinement.
• Deep learning model (Transformer and CNN) for diagnostic prediction.
• Functional annotation and pathway analysis.

Main Results:

• A panel of 220 methylation sites showed stable diagnostic performance in an independent cohort.
• Identified established OA genes (e.g., TGFBR2, SMAD3, PPARG, MAPK3).
• Suggested INHBB, AMH, and INHBE as potential novel effector genes in OA.

Conclusions:

• A robust, methylation-based framework for OA biomarker discovery has been established.
• The study provides new insights into the epigenetic mechanisms underlying OA.
• The identified biomarkers hold potential for early OA diagnosis and therapeutic strategies.