Decoding Promoter Activity from DNA Sequence using Pre-trained Language Models
The paper trains a transformer-based model on synthetic Drosophila promoters to predict promoter activity and shows it achieves high accuracy with sequence alone and improves with biological context.