Genetic architecture and genomic prediction accuracy of apple quantitative traits across environments

Abstract

Implementation of genomic tools is desirable to increase the efficiency of apple breeding. The apple reference population (apple REFPOP) proved useful for rediscovering loci, estimating genomic prediction accuracy, and studying genotype by environment interactions (G×E). Here we show contrasting genetic architecture and genomic prediction accuracies for 30 quantitative traits across up to six European locations using the apple REFPOP. A total of 59 stable and 277 location-specific associations were found using GWAS, 69.2% of which are novel when compared with 41 reviewed publications. Average genomic prediction accuracies of 0.18–0.88 were estimated using single-environment univariate, single-environment multivariate, multi-environment univariate, and multi-environment multivariate models. The G×E accounted for up to 24% of the phenotypic variability. This most comprehensive genomic study in apple in terms of trait-environment combinations provided knowledge of trait biology and prediction models that can be readily applied for marker-assisted or genomic selection, thus facilitating increased breeding efficiency.

Combining genetic resources and elite material populations to improve the accuracy of genomic prediction in apple

“…Second, if some alleles linked to the desired trait are in low frequency in one population and segregate in the other population, using only the population with the low-frequency alleles as the training set will lead to incorrect marker effect estimations. For example, Migicovsky et al (2021) showed that the NAC18.1 gene marker, which is linked to harvest date and fruit firmness, is homozygous for the favorable allele in the nine most marketed varieties in the United States, and several marker alleles detected by GWAS in the REFPOP panel are fixed in the elite population whereas they segregate in the genetic resources ( Jung et al , 2021 ). As expected in this case, the prediction for harvest date in the hybrids when using only the elite material in the training set led to predictive abilities lower than when using both populations.…”

“…As expected in this case, the prediction for harvest date in the hybrids when using only the elite material in the training set led to predictive abilities lower than when using both populations. Another such example would be the Rvi6 gene that confers resistance to apple scab and all QTLs in the associated introgressed segment: the favorable allele at Rvi6 segregates in elite material ( Jung et al 2021 ) but is absent in old varieties, because the gene introgression from the wild relative M. floribunda is recent ( Gessler and Pertot 2012 ).…”

“…Traits related to yield, fruit quality and phenology were measured between 2018 and 2020 using a common protocol in each location. The detailed protocol is described in Jung et al ( 2021). Harvest date and fruit over-color were the only two traits that were measured in both the FBo-Hi and the REFPOP panels.…”