Browsing by Author "Saludares, Rica"

Filter results by typing the first few letters
Now showing 1 - 1 of 1
  • No Thumbnail Available
    Item
    A multi-omics multi-environment prediction in pulse crop
    (North Dakota State University, 2024) Saludares, Rica
    Understanding the genetic bases underlying seed yield and protein, and eventually recombining them in desired genetic backgrounds, continues to be a challenge to pulse crop breeders. Phenotypic selection for seed yield and protein in preliminary yield trials is hindered by the need to phenotype a large number of early-generation lines (>10,000) with limited seeds, resulting to trials with few replications and limited environments. In this study, we evaluated and applied a multi-trait multi-environment (MTME) and a multi-omics prediction framework to address phenotyping bottleneck and the complexities underlying negatively correlated traits, and maximize connectivity among genotypes for predicting performance of untested genotypes in diverse set of environments. Using over 200 NDSU modern advanced breeding lines and 300 USDA diverse accessions, our findings demonstrated that MTME prediction significantly enhanced predictive ability by 1.3 and 1.8-fold for yield and protein, respectively. For the environments with low heritability of tested trait, however, using the MTME prediction led to small increases in prediction accuracy. To further maximize connectivity among genotypes and environments, a subset of individuals was included from the testing population that led to 1.6 and 1.2-fold improvement for yield and protein, respectively. Incorporating additional orthogonal information such as gene expression (RNA) into the prediction framework showed potential for further increasing prediction accuracy. Using ~300 USDA diverse accessions assessed in two environments, integrating genotypic and expression data (DNA+RNA) resulted to higher predictive ability (0.48-0.55) over using DNA only (0.42) or RNA only (0.43-0.53). Overall, we found that maximizing the relationship among genotypes and environments, along with integration of additional orthogonal information (e.g. RNA) into genomic prediction framework can further enhance predicting performance of untested genotypes in diverse environments.