Optimizing whole-genomic prediction for autotetraploid blueberry breeding

Abstract

Blueberry (Vaccinium spp.) is an important autopolyploid crop with significant benefits for human health. Apart from its genetic complexity, the feasibility of genomic prediction has been proven for blueberry, enabling a reduction in the breeding cycle time and increasing genetic gain. However, as for other polyploid crops, sequencing costs still hinder the implementation of genome-based breeding methods for blueberry. This motivated us to evaluate the effect of training population sizes and composition, as well as the impact of marker density and sequencing depth on phenotype prediction for the species. For this, data from a large real breeding population of 1804 individuals were used. Genotypic data from 86,930 markers and three traits with different genetic architecture (fruit firmness, fruit weight, and total yield) were evaluated. Herein, we suggested that marker density, sequencing depth, and training population size can be substantially reduced with no significant impact on model accuracy. Our results can help guide decisions toward resource allocation (e.g., genotyping and phenotyping) in order to maximize prediction accuracy. These findings have the potential to allow for a faster and more accurate release of varieties with a substantial reduction of resources for the application of genomic prediction in blueberry. We anticipate that the benefits and pipeline described in our study can be applied to optimize genomic prediction for other diploid and polyploid species.

Genomic prediction in an outcrossing and autotetraploid fruit crop: lessons from blueberry breeding

“…In a recent study, we demonstrated that such numbers are quite conservative for genomic prediction. By combining a simple genetic parametrization (ratio) and low-to-mid sequencing depth (6x-12x), we achieved similar predictive accuracies as higher-depths for blueberry traits with different genetic architectures (de Bem Oliveira et al, 2020). Similar results are also reported by Zheng et al (2020).…”

“…About half of the capture-seq genotyping probes that were originally developed based on a draft genome assembly were discarded afterwards based on the high-quality genome, without compromising genetic association and ge nomic prediction analyses (Benevenuto et al , 2019). We also explored additional optimizations to reduce costs, regarding the number of individuals per family, number of markers, and sequencing depth (de Bem Oliveira et al , 2020). Moreover, new genomics methods and tools have been developed in the last decade for the polyploid community, including allele dosage estimation, haplotype reconstruction, and the use of different relationship matrices (Bourke et al , 2018).…”

“…When confronting the problem of modeling the relationship between molecular markers and variation in the observed traits, an important question to keep in mind is what statistical method could better describe this relationship (FerrÃo et al , 2019). In recent years, we have investigated statistical and biological aspects underlying the implementation of genomic prediction in autopolyploid species, including (i) the importance of accounting for allele dosages in whole-genome statistical models (de Bem Oliveira et al , 2019); (ii) the relevance of multiple gene actions, including additive and non-additive genetic sources (Amadeu et al , 2020; Zingaretti et al , 2020); and finally, (iii) the impact of sequencing depth of coverage, when sequence-based genotyping approaches are used (de Bem Oliveira et al , 2020).…”

Genomic Prediction in Family Bulks Using Different Traits and Cross-Validations in Pine

“…Genomic (Elshire et al 2011), statistical (Meuwissen et al, 2001; Gianola et al, 2009), and computational advances have allowed significant increases in genetic gain by applying genomic prediction (GP) in breeding programs across several species (e.g., Hayes et al, 2009; Fe et al, 2015, 2016; Gezan et al, 2017; de Bem Oliveira et al, 2020; Amadeu et al, 2020). Taking advantage of the ever-reducing cost of molecular markers (Wetterstrand, 2020), the concept of GP was derived (Meuwissen et al, 2001) as an alternative method to marker-assisted selection (MAS).…”

“…Pioneer studies implementing GP in plants were performed in mayor crop species with traditional hybrid selection such as maize (Massman et al, 2013; Combs and Bernardo, 2013) and trees (Resende et al, 2012; Kumar et al, 2012), or variety selection in self-pollinating species (Poland et al, 2012). Genomic prediction showed to be a powerful tool to achieve higher genetic gain in plant breeding in many other species (Crossa et al, 2017; Lara et al, 2019; de Bem Oliveira et al, 2020; Esfandyari et al, 2020). Large commercial breeding companies have been applying GP; however, the success of the process depends strongly on the species architecture and the breeding program scheme (Xu et al, 2020; Voss-Fels et al, 2019)…”