AI Maps 3D Super-Enhancers, IDs Cell Regulators

New algorithm reveals how gene expression is controlled in 3D protein condensates

Published on Mar. 10, 2026

Got story updates? Submit your updates here. ›

Scientists at St. Jude Children's Research Hospital have developed a new machine learning algorithm called BOUQUET that can identify 3D "super-enhancers" - groups of enhancers and their associated proteins that regulate gene expression in cells. The research shows that these 3D super-enhancer communities are linked to transcriptional protein condensates, high-density membraneless droplets in cell nuclei that control specialized cell identities. The findings provide new insights into how cells regulate genes and could have implications for understanding disease-causing gene expression.

Why it matters

Understanding how gene expression is regulated in the 3D environment of the cell nucleus is crucial for unraveling the molecular mechanisms that control cell identity and function. This research provides a new computational tool to map these 3D regulatory networks, which could lead to better understanding of how dysregulated transcription contributes to diseases like cancer.

The details

The BOUQUET algorithm uses machine learning and graph theory to identify groups of enhancers and their associated proteins that interact in 3D space to control gene expression. The researchers found that the enhancer communities with the highest levels of associated proteins, dubbed "3D super-enhancers", correspond to the locations of transcriptional protein condensates in the nucleus. They were able to predict and confirm a new gene that interacts with these condensates, and observed two genes from the same 3D super-enhancer community being co-transcribed within the same condensate.

  • The study was published on March 9, 2026.

The players

Brian Abraham

An assistant professor in the Department of Computational Biology at St. Jude Children's Research Hospital and the corresponding author of the study.

Kelsey Maher

A co-first author of the study and a postdoctoral researcher in the Department of Computational Biology at St. Jude.

Jie Lu

A co-first author of the study and a postdoctoral researcher in the Department of Computational Biology at St. Jude.

St. Jude Children's Research Hospital

A pediatric treatment and research facility located in Memphis, Tennessee that focuses on catastrophic childhood diseases.

BOUQUET

A new machine learning algorithm developed by researchers at St. Jude to map 3D enhancer-gene regulatory networks in cells.

Got photos? Submit your photos here. ›

What they’re saying

“With BOUQUET, we can quantify the activating protein apparatus that is associated with each gene. This assignment gave us two major advances: predicting gene expression from protein binding maps and finding which genes are likely interacting with transcriptional condensates.”

— Brian Abraham, Assistant Professor, Department of Computational Biology, St. Jude Children's Research Hospital (Mirage News)

“The data argue that communities are fundamental units of gene regulation because their parts show correlated activities, and perturbations made to one part of the community affect the whole community.”

— Jie Lu, Postdoctoral Researcher, Department of Computational Biology, St. Jude Children's Research Hospital (Mirage News)

What’s next

The researchers plan to further investigate how the 3D super-enhancer communities and associated transcriptional condensates may contribute to dysregulated gene expression in diseases like cancer.

The takeaway

This new computational approach to mapping 3D gene regulatory networks provides valuable insights into the fundamental units of transcriptional control in cells, with potential applications for understanding the molecular basis of diseases driven by aberrant gene expression.