Editing a downstream repressive cis-regulatory element of TAW1 unlocks yield potential in rice

To systematically dissect the direction and magnitude of cis-regulatory sequence effects on gene transcription in rice, we integrated ATAC-seq and RNA-seq data from 22 tissues and organs of the japonica rice cultivar Nipponbare and developed RiceTissueX, a deep learning framework composed of three functionally complementary submodels. RiceTissueX jointly predicts the regulatory direction (activation or repression), effect size, and tissue specificity of candidate cis-regulatory sequences, and further supports in silico perturbation analysis of regulatory elements, providing computational guidance for precise genome editing targeting cis-regulatory elements.This is a preprint version of the manuscript. The work has not yet been peer-reviewed.