Multi-omics integration maps CHH methylation and gene regulatory networks across heat, drought, and salt stress in rice.

Plant phenotypes arise from complex interactions among genetic, epigenetic, and environmental factors. Deciphering the molecular mechanisms responsible for phenotypic variation requires comprehensive multi-omics approaches that integrate diverse genomic, transcriptomic, and environmental datasets. Here, we perform high-throughput phenotyping of rice subspecies cultivars (Nip and 93-11), covering key growth stages under heat, drought, and salt stress. Using the EfficientNet framework, we generate image embeddings for cultivar classification and stress-type identification. Whole-genome bisulfite sequencing (WGBS) identifies elevated CHH methylation under heat and drought stresses but reduced levels under salt stress at ripening. Furthermore, we construct a multi-omics network and uncover heat-responsive subnetworks. Within this, OsCam3 emerges as a key regulator of heat stress responses in both Nip and 93-11. Additionally, OsCam3 underwent different selection in the japonica and indica subpopulations. Altogether, our study highlights the interplay among epigenetic, transcriptional, and phenotypic factors in shaping stress responses, providing valuable genetic resources for future climate-resilient rice breeding.