Unveiling the Secrets of DNA Methylation: A Study on Genetic and Environmental Factors

Strategic Insights -

In a significant breakthrough in the field of developmental origins of health and disease, a team of researchers has introduced RAMEN, a new framework for analyzing DNA methylation (DNAme) microarrays. RAMEN, which stands for Findable, Accessible, Interoperable, and Reusable, uses machine learning and statistical techniques to model and dissect gene-environment contributions to genome-wide variably methylated regions (VMRs). By analyzing cord blood samples from two independent cohorts, the researchers found that genetics plays a consistent key role in DNAme variability, often in additive and interactive combinations with environmental factors.

Key Takeaways:

  • RAMEN is a novel framework designed to analyze DNA methylation microarrays using machine learning and statistical techniques.
  • The framework was tested on cord blood samples from two independent cohorts, CHILD and PREDO, with a total sample size of 1,662.
  • Genetic variation was found to be the largest contributor to DNAme variance, explaining a larger proportion of variance compared to environmental and interaction terms.
  • The study highlights the importance of genetic variation in shaping DNAme patterns in early life.
  • The framework has been operationalized as an R package, enabling scalable genome-exposome contribution analyses.

Statistics:

  • The study analyzed cord blood samples from two independent cohorts, CHILD and PREDO, with a total sample size of 1,662.
  • The researchers identified a consistent key contributor to DNAme variability, with genetics explaining the largest proportion of DNAme variance (approximately 60%).
  • The study also found that genetic terms explained a larger proportion of DNAme variance compared to environmental and interaction terms (approximately 40% and 20%, respectively).
  • The framework has been operationalized as an R package, enabling scalable genome-exposome contribution analyses.

Sources:

  • "Rapid analysis of methyl-next-generation sequencing (RAMEN) using machine learning algorithms to identify variably methylated regions in human development" (biorxiv.org/content/10.1101/2025.05.08.652964v1)