Cross-platform DNA methylation classifier for the 8 molecular subtypes of group 3/4 medulloblastoma

The discovery of molecular subgroups of medulloblastoma, a malignant pediatric brain cancer, has enabled personalized treatment strategies. In 2019, a study formed a consensus on 8 novel subtypes within Groups 3 and 4, each displaying heterogeneous characteristics. Classifiers are essential for translating these findings into clinical practice by supporting clinical trials, personalized therapy development and application, and patient monitoring. This study introduces the first machine learning classifier capable of distinguishing the 8 consensus-defined Group 3 and Group 4 medulloblastoma subtypes across both Illumina 450K and EPIC (850K) arrays. Across two independent test sets, the model achieved a weighted F1 score of 0.95 and a balanced accuracy of 0.957, demonstrating consistent performance across both array platforms. We also developed and deployed a web application capable of classifying samples using 450K or EPIC methylation data. This tool provides subgroup predictions with associated probability reporting. As a robust solution, it ensures backward compatibility with existing platforms while extending applicability to newer methylation array technologies, significantly enhancing accessibility for researchers. Overall, this work advances precision medicine with potential to improve clinical outcomes for patients within the most prevalent medulloblastoma subgroups (Group 3 and Group 4).