Genomic Selection for Fruit Weight and Soluble Solid Contents in Peach

Fruit weight and soluble solid content (SSC) are two important quantitative traits in peach which are of importance to breeders. However, performing early prediction using a single marker is challenging as the traits are controlled by multiple minor genes. Genomic selection, a novel genome-wide tool, has been applied in fruit crops and can potentially enhance the breeding efficiency of these quantitative traits. However, its effects in peach and influencing factors require further investigation.

Establish a whole-genome selection technology system for peach single fruit weight and SSC, and laid a methodological foundation for the establishment of efficient molecular breeding technology system for peach.

The objectives of this study were to assess the accuracy of prediction of peach fruit weight and SSC in natural and hybrid populations using genomic selection. Here, a training population of 520 individuals was selected. Using genotypic data for 48 398 single nucleotide polymorphisms (SNPs) obtained from the resequencing results of the above training population, a total of 11 genome-wide prediction models were built to select the optimum model for fruit weight and SSC. Subsequently, the genomic breeding values of a small natural population of 56 individuals and 29 hybrid populations comprising a total of 1 145 seedlings were calculated.

The average sequencing data of each variety of the three groups was 1.95-3.52 Gb, and the sequencing depth was 5.29-10.79×. The sequencing data of the training natural population was aligned with the reference genome, and a total of 5 065 726 SNPs were obtained. After removing the SNPs with a high missing rate (>20%) and minor allele frequency of <0.05, a total of 48 398 SNPs on the genome were randomly selected for constructing whole-genome selection models for the training population. The models with the highest prediction accuracy for fruit weight and SSC were BayesA and randomforest, respectively. Using the above two models, it was found that the goodness of fit between the predicted breeding values and observed phenotype of fruit weight was 0.4767-0.6141, which was higher than that of SSC (0.3220-0.4329) in the natural populations. In hybrid populations, the prediction accuracy of fruit weight was 0.2319-0.4870, which was also higher than that of SSC (0.0200-0.2793). The results also showed that the prediction model constructed by training natural populations was more accurate in predicting natural populations than hybrid populations. Taking fruit weight as an example, it was also found that only 17.78% of the seedlings needed to be retained by genomic selection when targeting large fruit. Genomic selection was significantly more efficient than single and double marker selection. Furthermore, the effects of population dispersion, heritability and population structure on prediction accuracy are also discussed. The results indicated that prediction accuracy may vary and be affected by a combination of several factors.

In this study, a suitable genomic selection model for peach fruit weight and SSC was screened, and it was confirmed that the prediction efficiency of genomic selection was significantly higher than that of single marker selection. The results indicated the potential of genomic prediction in accelerating breeding progress of these two quantitative traits in peach.