Modeling gene regulation across cellular contexts with deep learning

Understanding how cells regulate gene expression is a longstanding goal of biology. Deep learning has advanced this effort through sequence-to-function models that predict functional genomic measurements from DNA and cell state models that learn patterns of coordinated gene expression. Connecting these approaches could help predict how sequence effects depend on the cellular environment and how regulatory sequence shapes responses to regulator perturbations. Here, we review emerging approaches to joint modeling and examine how biological measurements, model design, and training data shape the regulatory relationships they can learn. We discuss evaluations that test generalization to unmeasured sequences and cellular conditions, alongside interpretation methods that generate mechanistic hypotheses. We then consider how incorporating cellular dynamics could help predict the propagation and persistence of regulatory responses. Finally, we discuss how pilot scaling studies could guide data acquisition and how AI agents could support model development, interpretation, and experimental design.

Authors

Institutions

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-19
DOI
https://doi.org/10.5281/zenodo.24374086
Primary Topic
Gene Regulatory Network Analysis
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Modeling gene regulation across cellular contexts with deep learning

Wunna Kyaw, Emily S. W. Wong, Maria Brbić, Ekin Deniz Aksu et al.
Zenodo (CERN European Organization for Nuclear Research)
Gene Regulatory Network Analysis
article

Modeling gene regulation across cellular contexts with deep learning

Wunna Kyaw, Emily S. W. Wong, Maria Brbić, Ekin Deniz Aksu, Alan E Murphy, Peter K. Koo, Ruchir Rastogi, Anshul Bharat Kundaje, Selin Jessa
article en

Abstract

Understanding how cells regulate gene expression is a longstanding goal of biology. Deep learning has advanced this effort through sequence-to-function models that predict functional genomic measurements from DNA and cell state models that learn patterns of coordinated gene expression. Connecting these approaches could help predict how sequence effects depend on the cellular environment and how regulatory sequence shapes responses to regulator perturbations. Here, we review emerging approaches to joint modeling and examine how biological measurements, model design, and training data shape the regulatory relationships they can learn. We discuss evaluations that test generalization to unmeasured sequences and cellular conditions, alongside interpretation methods that generate mechanistic hypotheses. We then consider how incorporating cellular dynamics could help predict the propagation and persistence of regulatory responses. Finally, we discuss how pilot scaling studies could guide data acquisition and how AI agents could support model development, interpretation, and experimental design.

Zenodo (CERN European Organization for Nuclear Research)
SIB Swiss Institute of Bioinformatics (CH), Mount Sinai Hospital (CA), University of Toronto (CA), Victor Chang Cardiac Research Institute (AU), Max Delbrück Center (DE), Lunenfeld-Tanenbaum Research Institute (CA), Cold Spring Harbor Laboratory (US), UNSW Sydney (AU), Max Planck Institute for Molecular Genetics (DE), École Polytechnique Fédérale de Lausanne (CH), Stanford University (US)
Gene Regulatory Network Analysis
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon