Deep learning the cis-regulatory code for gene expression in selected model plants
Problem
Understanding how non-coding regulatory DNA sequences determine gene expression levels in four plant species, including linking genetic variation to expression changes.
Why it matters
Gene regulation underlies crop traits and genetic variation, so decoding it aids plant breeding and genetics.
What the AI did
Trained interpretable deep learning models to predict gene expression profiles from gene flanking regions, enabling feature selection and cross-species regulatory analysis.
Results
Over 80% prediction accuracy; cross-species identification of conserved and species-specific regulatory features; causal links revealed across fourteen tomato genomes.
Impact
The models predicted genotype-specific expression and identified regulatory features, but the full cis-regulatory code is not claimed as solved.
Record
| Venue | Nature Communications |
|---|---|
| Published | 2024 |
| Work type | article |
| Authors | Fritz Forbang Peleke, Simon Maria Zumkeller, Mehmet Gültaş, Armin Otto Schmitt, Jędrzej Szymański |
| Citations | 86 (OpenAlex, at import time) |
| References | 84 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1038/s41467-024-47744-0 |
| Open access | gold |
| Keywords | deep learning, cis-regulatory, gene expression, plant genomics, UTR |
Importance rating
significant — thousands to millions affected
Linking regulatory DNA to expression underlies crop traits and genetic variation across four plant species including crops; abstract implies breeding relevance but states no scale.