Predicting transcriptional regulators in plants in the era of artificial intelligence

Abstract

Identifying transcriptional regulators that control important biological pathways in plants is fundamental to understanding regulatory mechanisms, network hierarchy, and phenotypic variation. This remains challenging because transcription factors (TFs) and their targets operate within highly interconnected, dynamic, and often redundant regulatory networks. Over the past two decades, advances in omics technologies, sequencing data generation, and computational tools have shifted gene discovery from single-gene studies to network-level investigation. At the same time, these advances have created a new challenge: how to extract biologically meaningful regulatory relationships from increasingly complex and high-dimensional datasets. Recent progress in multi-omics integration, machine learning, deep learning, and emerging foundation-model approaches is beginning to address this challenge and is reshaping how transcriptional regulators, targets, and regulatory relationships are predicted in plants. In this review, we summarize advances in network-enabled gene discovery, discuss how multi-omics and AI are transforming transcriptional target prediction, and consider how these developments may lead to predictive models of plant gene regulation with applications in crop improvement and synthetic biology.