# Methylation classifier for brain tumours

Source: https://onco.cc/technologies/cns-tumour-methylation-classifier/  
OnCo record `cns-tumour-methylation-classifier` (Technology). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

A test that reads the methylation pattern of a brain tumour's DNA and matches it to a reference library, giving a more reliable diagnosis than the microscope alone.

## Summary

Brain tumour diagnosis by microscopy is notoriously inconsistent. In 2018 the Heidelberg group published a machine-learning classifier that assigns tumours to one of more than eighty methylation classes from a genome-wide methylation array, reclassifying a meaningful share of cases and revealing new entities. The approach is now embedded in the WHO classification of central nervous system tumours and used in reference neuropathology centres.

## Fields

- Kind: Technology
- Status: established
- Last checked: 2026-09-04
- Principle: DNA methylation is measured on an array, and a random-forest classifier trained on thousands of reference tumours returns the best-matching class with a calibrated score.
- Since: 2018
- Strengths: Reproducible, objective diagnosis; Detects rare and new entities; Works on small or archived samples
- Limitations: Array cost and turnaround; Unclassifiable cases remain; Requires reference-centre expertise

## Sources

- Wikipedia: https://en.wikipedia.org/wiki/DNA_methylation
- Wikipedia: https://en.wikipedia.org/wiki/DNA_methylation

## Connected records

- cancers: [Brain and spinal cord tumours (all types)](https://onco.cc/cancers/brain-tumours/), [Childhood cancers (all types)](https://onco.cc/cancers/childhood-cancers/), [Glioma & glioblastoma](https://onco.cc/cancers/glioblastoma/)
- fronts: [Diagnostics & Biomarkers](https://onco.cc/fronts/diagnostics/)
- technologies: [DNA methylation profiling](https://onco.cc/technologies/methylation-profiling/)

---
JSON: https://onco.cc/api/v1/entities/cns-tumour-methylation-classifier.json