Year: 2026
Type: preprint
Abstract: Selection and mate allocation shape performance and diversity in future breeding populations. Although numerous criteria for evaluating and selecting crosses exist, few tools offer a unified, scalable framework for different breeding schemes, multi-trait objectives, and diversity management. We present CrossingTools, an open source R package for predicting cross performance and optimising crossing plans using marker genotypes and genetic effects. Crosses can be evaluated based on their expected progeny mean, optimal haploid value, or superior progeny value using segregation (co)variance models for inbred and heterozygous parents with additive and dominance effects. Multi-trait objectives can be defined using economic weight or desired gains indices, while crossing plans can be optimised to balance predicted performance and genomic diversity through optimal cross selection. Efficient C++ subroutines enable large-scale cross evaluation, allowing the screening of entire breeding populations. To demonstrate its application, we simulated a line breeding programme with correlated traits and found that, under genomic selection, superior progeny value and optimal haploid value improved upper-tail progeny performance more strongly than population mean gain and preserved more genetic variance than selection based on expected cross means. The criteria also affected genetic correlations differently as superior progeny value tended to shift correlations in directions favourable to the selection index, optimal haploid value largely maintained correlations, and selection on expected cross means often produced less favourable changes. Optimal cross selection reduced variance depletion at the cost of lower short-term gain. CrossingTools thus offers a framework for integrating progeny prediction, multi-trait selection, and genetic diversity management in plant breeding.
Source: Research Square
Language: English
Subfield: Genetics
Domain: Life Sciences
SDG: Zero hunger
Open Access status: green
OpenAlex ID: W7236689705
