MethylGPT

01 Sep 2024 in Research on Foundation-models, Biology, Methylation Last modified at: 2026-02-14

A foundation model approach to understanding human methylation patterns - under review at Nature Methods.

Overview

MethylGPT is a foundational GPT-like model designed for human methylation data, applying transformer architectures to biological sequence analysis.

Background

DNA methylation is an epigenetic modification that plays a crucial role in gene regulation, development, and disease. Traditional analysis methods are often limited in their ability to capture complex patterns across the methylome.

Approach

MethylGPT applies the foundation model paradigm—successful in natural language processing—to methylation data:

• Pre-training on large-scale methylation datasets
• Transfer learning to downstream tasks
• Interpretable representations of methylation patterns

Key Contributions

1. Scalable Architecture: GPT-style transformer adapted for methylation arrays
2. Pre-trained Foundation Model: Can be fine-tuned for various downstream tasks
3. Biological Insights: Learned representations capture meaningful biological signals

Status

Currently under review at Nature Methods.

Citation

@article{ying2024methylgpt,
  title={MethylGPT: Foundational GPT-like Model for Human Methylation Data},
  author={Ying, Alex and Song, Jinyeop and Cui, Haoran and others},
  journal={bioRxiv},
  year={2024}
}

Preprint: bioRxiv