New ML Method Boosts CNS Tumor Classification Accuracy

Researchers have developed a new, methodologically rigorous machine learning technique for classifying Central Nervous System (CNS) tumors based on DNA methylation profiles. This innovative approach integrates Sparse Random Projection for efficient dimensionality reduction with multinomial logistic regression for robust classification. The study rigorously evaluated the new method against a widely used reference classifier. On a 2,801-sample reference cohort, the proposed model achieved an impressive mean accuracy of 96% under stratified 3-fold cross-validation. Crucially, when tested on an independent 1,104-sample clinical evaluation cohort, the method reached 86% accuracy at the 91-class level and 93% at the methylation class family level. These figures represent a significant improvement of approximately 4 to 5 percentage points over the previous state-of-the-art, offering clinically relevant gains that can directly impact cancer subtype assignment and treatment decisions.