Skip to main content
Videos

AI-Driven DNA Methylation Profiling Advances Tumor Classification and Diagnosis


Clinical Summary: 

  • Kenneth Aldape, MD, examines advances in tumor classification using DNA methylation profiling and artificial intelligence (AI), including the development of histopathology-based models that infer epigenomic and molecular characteristics from routine tissue images.
  • DNA methylation classifiers improve the distinction between morphologically similar  tumor entities, while AI models trained on hematoxylin and eosin images can predict methylation patterns and classify multiple tumor types. In a comparative evaluation, AI-assisted diagnosis improved classification performance among participating pathologists.
  • These findings highlight the potential for AI-driven epigenomic analysis to improve diagnostic accuracy, expand access to molecular classification, and support pathologists in challenging cases, while further validation is needed before broader clinical implementation.

Kenneth Aldape, MD, Mayo Clinic, Rochester, Minnesota, discusses advances in tumor classification using DNA methylation profiling and artificial intelligence (AI), including the development of computational models that infer molecular characteristics from routine histopathology images.

Dr Aldape reviews how DNA methylation signatures can distinguish morphologically similar tumors and identify previously unrecognized tumor entities. He also discusses AI models that predict epigenomic profiles and gene expression from hematoxylin and eosin images, highlighting findings demonstrating improved diagnostic performance when pathologists incorporated AI-generated predictions into their assessments.


Source: 

Aldape K. DNA methylation and artificial intelligence for the diagnosis and classification of CNS tumors. Presented at the AMP Annual Meeting and Expo. November 10-14, 2026; Seattle, Washington.

© 2026 HMP Global. All Rights Reserved.
Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of Oncology Learning Network or HMP Global, their employees, and affiliates.