NB-TeaBase: A Multi-Omics Database and Genomic Selection Platform for Robust-Bud Tea Breeding

Tea breeding increasingly requires a continuous evidence chain from germplasm identity to molecular variation and progeny performance, but cultivar passports, genome resources, transcriptome profiles, metabolite data and breeding records are often managed separately. Here, we present Ningbo TeaBase (NB-TeaBase), an open, cultivar-centered database for robust-bud tea (Camellia sinensis) breeding. The platform uses named cultivars as the organizing unit and integrates germplasm passport information for eight elite cultivars with 23 chromosome-level genomes, whole-genome resequencing variants, RNA-seq expression profiles, untargeted LC-MS/MS metabolite features, annotations and downloadable records. Shared cultivar, gene, variant, pathway and metabolite identifiers allow users to characterize a cultivar across genomic, transcriptomic and metabolic layers and to connect these profiles with breeding records. A parental-prediction module incorporates genome-wide markers and phenotypes from 122 progeny assigned to seven observed parental combinations; it reports genomic estimated breeding values (GEBVs) for candidate crosses and observed-family general and specific combining-ability summaries for seven traits. By combining germplasm documentation, multi-omics evidence and progeny-informed cross ranking, NB-TeaBase supports cultivar evaluation, parent selection, cross prioritization and breeding strategy formulation. Current prediction outputs remain exploratory because the phenotypic training data cover only seven parental combinations and require broader validation.