Corresponding author: s.dreisigacker@cgiar.org

Over the past three decades, the International Maize and Wheat Improvement Center (CIMMYT) has let the way in developing and utilizing synthetic hexaploid wheat (SHW) to facilitate the transfer of genes from Ae. tauschii and durum wheat to bread wheat. Despite the generation and characterization of numerous SHWs, accurately forecasting their phenotypic performance has proven challenging, especially when relying solely on the performance of the parental lines.

To address this challenge, we investigated methods originally designed for predicting traits in intraspecific hybrids and applied them to the context of allopolyploids, specifically the cross of durum wheat with Ae. tauschii. We successfully predicted the phenotypes of SHWs for three significant quantitatively inherited diseases in wheat: tan spot, Septoria nodorum blotch and spot blotch.

Genomic prediction models, incorporating both pedigree information and non-linear kernels, surpassed the predictive capabilities of models relying solely on linear kernels. Consequently, we have initiated the implementation of recurrent, early-generation genomic selection for crosses between SHW and elite spring wheat lines from CIMMYT. This strategic deployment aims to generate multiple layers of pre-breeding populations and novel candidate lines, enriched with novel allelic variations from the SHW crucial for enhancing quantitative disease resistance.