The differentiation of a totipotent cell into a somatic cell type is a highly orchestrated process characterized by changes in the epigenetic landscape of a cell shaping its transcriptomic identity. One of these epigenetic layers is DNA methylation, which is mostly restricted to the symmetrical CpG context and about 60-80% of the approximately 28 million CpGs in the human genome are typically methylated. However, virtually all cancers deviate from this canonical methylation profile in often very specific manner. Here, we utilized a large, previously published cohort of Illumina methylation arrays (Capper et al., 2018) to train a real-time feasible machine-learning model together with rapid Nanopore sequencing to enable intra-operative brain cancer classification.