Consolidating 23 years of historical data from an irrigated subtropical rice breeding program in Uruguay

Breeding programs generate vast amount of data which are often scattered in separate files. This hinders the application of modern breeding tools such as multi-environment analyses and genomic selection. This research work describes the process of consolidating 23 years of phenotypic, pedigree, and genomic records from the Uruguayan national rice (Oryza sativa L.) breeding program, and the features and structure of the resulting database. Using a custom-made R code, we gathered all the available data from 1997 to 2020 corresponding to field trials, blast disease evaluation nurseries, laboratory analyses of milling and cooking quality, pedigree information, and genomic information for selected advanced breeding lines, and organized it into a relational database. Records of 996 trials in 12 locations over a span of 23 years, 91,636 field plots with information on 14 phenotypic variables, pedigree for 19,447 genotypes, and genomic information regarding 61,260 single nucleotide polymorphism (SNP) markers for 965 genotypes were recovered. The dataset is structured in trials, phenotypes, lines, genomic information, and SNP tables in an easy-to-access relational database, which will be a valuable resource for rice breeding.