Identification of Reliable Biomarkers for ALS Through Machine Learning Approach - PubMed
3 hours ago
- #Biomarkers
- #Machine Learning
- #Amyotrophic Lateral Sclerosis
- Machine learning (ML) methods, specifically XGBoost and random forest, were used to identify potential biomarkers for Amyotrophic Lateral Sclerosis (ALS).
- The study utilized RNA-Seq data from motor neuron disease patients and healthy controls, processed through a five-fold cross-validation framework.
- Both ML models demonstrated high classification performance, with random forest achieving 98.8% accuracy and XGBoost 97.6% accuracy.
- The Decorin (DCN) gene consistently ranked in the top 10 features across both models, highlighting its stability and potential as a reliable biomarker.
- The research emphasizes the effectiveness of combining transcriptomic data with machine learning to discover key genes for ALS diagnosis and therapy.