Official websites use .gov

A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS

A lock ( 🔒 ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Need Help

Integrating Language Models and Topology for Multi-Scale R-Loop Prediction

Award Number: R01GM169397
ORGANIZATION: NATIONAL INSTITUTE OF GENERAL MEDICAL SCIENCES
OPDIV: NIH
AWARD CLASS: DISCRETIONARY
AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)
PERIOD OF PERFORMANCE START DATE: 09/02/2026
PERIOD OF PERFORMANCE END DATE: 08/31/2029

Group Awards By:

View Award Description

Integrating Language Models and Topology for Multi-Scale R-Loop Prediction - PROJECT SUMMARY (See instructions): R-loops are three-stranded structures composed of an RNA:DNA hybrid and a displaced single DNA strand that typically form during transcription and constitute 3-5% of the genome across taxa. While R-loops are fundamental to genome regulation, aberrant R-loop formation, often driven by topological stress, promotes replication stress and double-strand breaks that drive chromosomal rearrangements. The physical and topological factors that drive R-loop formation, stability, and complex 3D architecture remain poorly understood. This research establishes a predictive, multi-scale, quantitative framework to analyze these factors by integrating experimental techniques with new mathematical and machine learning models. Specific Aim 1 integrates single-molecule data with probabilistic grammars and Large Language Models (LLMs) to create a new class of predictive models of R-loop formation on natural genomic sequences. Specific Aim 2 determines how DNA topology regulates the formation and 3D structure of R-loops by coupling biochemical experiments and mathematical approaches based on topology and discrete mathematics. The project uses single-molecule footprinting, DNA topology experiments, and Atomic Force Microscopy to quantify the effect of topology and address fundamental questions about R-loop entanglement. The research introduces a probabilistic grammar framework, a structure based on tile systems, and topological models rooted in lattice walks and polygons to formalize entanglement in multi-stranded nucleic acids. These models predict how and where R-loops initiate, and how topology and other factors shape 3D structure. The integration of machine learning with formal grammars will enable predictive models that capture both sequence-level signals and higher-order topological entanglement, providing a unified approach to describe, classify, and predict R-loop structures. By identifying hotspots for R-loops along human genomic sequences, this research provides the biophysical framework necessary to define the mechanistic link between R-loop formation, genome instability, and disease.


Issue Date FY	Funding FY	Legal Entity Name	Legal Entity Address	Legal Entity City	Legal Entity State	Legal Entity Zip Code	Legal Entity COUNTY	Legal Entity COUNTRY	Assistance Listing	Award Code	Budget Year	Action Date	Action Type	Action Amount

Issue Date FY: 2026 ( Subtotal = $409,731 )
2026	2026	UNIVERSITY OF CALIFORNIA, DAVIS	1850 RESEARCH PARK DR STE 300	DAVIS	CA	95618	YOLO	USA	Biomedical Research and Research Training	000	1	9/2/2026	NEW	$409,731
														Subtotal = $409,731

Grand Total All Awards = $409,731

Top

All Categories

About

Search

Reports

Data Submission

Award Information

Integrating Language Models and Topology for Multi-Scale R-Loop Prediction

Award Number: R01GM169397

ORGANIZATION: NATIONAL INSTITUTE OF GENERAL MEDICAL SCIENCES

OPDIV: NIH

AWARD CLASS: DISCRETIONARY

AWARD ACTIVITY TYPE: SCIENTIFIC/HEALTH RESEARCH (INCLUDES SURVEYS)

PERIOD OF PERFORMANCE START DATE: 09/02/2026

PERIOD OF PERFORMANCE END DATE: 08/31/2029

Federal Websites

Department of Health & Human Services

HHS Operating Divisions

HHS Staff Divisions

Download A Document Viewer