Comprehensive mapping and modelling of the rice regulome landscape unveils the regulatory architecture underlying complex traits

We systematically depicted chromatin accessibility profiles in various tissues across the life cycle of three representative rice cultivars using the UMI-ATAC-seq method16, a modified ATAC-seq protocol developed in our lab. From 145 ATAC-seq datasets, we obtained a total of 117,176 unique open chromatin regions (OCRs), accounting for ca 15% of the rice genome. By integration of RNA-seq data for matched tissues, we predicted putative robust target genes for regulatory elements based on the correlation of gene expression and flanking chromatin accessibility across all tissues. We further inferred tissue- or stage-specific regulatory networks through TF footprinting analysis, and identified cultivar-associated OCRs by comparing the regulatory landscapes between indica and japonica rice subspecies. Finally, our analysis revealed that trait-associated variants are preferentially located in tissue-specific OCRs, which enabled us to identify causal associations between 209 complex agronomic traits and noncoding regulatory variants using this OCR landscape. Overall, these data not only serve as a foundational resource for the plant research community but also provide valuable regulatory variants for molecular breeding.