Predictive Ability of Genomic Estimated Family Values (GEFV)

Genomic selection (GS) has been widely used to compute genomic estimated breeding values (GEBV) of individuals; however, it has only been applied to breeding cattle and major plant crops. Implementing GS in minor crops at the individual level is limited due to high costs. Additionally, in some crops selection is performed at the family level. We aimed to study the implementation of genome-wide family selection (GWFS) in a breeding population of loblolly pine (Pinus taeda L.) composed of 63 families (5-20 individuals per family), which was phenotyped for four traits (stem diameter, rust susceptibility, tree stiffness and lignin content) and genotyped using an Illumina Infinium assay with 4740 polymorphic SNPs. Phenotypic and genotypic data was pooled at the family level in silico. Phenotypes were averaged across replicates for all the individuals and allele frequency was computed for each SNP. Marker effects were estimated at the family (GEFV) and individual (GEBV) level with Bayes-B using the package BGLR in R and models were validated using 10-fold cross validations. Predicted ability was computed by correlating phenotypes with GEFV and GEBV. Predicted ability was always higher for GEFV (0.51±0.02, 0.50±0.03, 0.47±0.03 and 0.28±0.03, respectively as order above) than for GEBV (0.45±0.02, 0.31±0.02, 0.39±0.03 and 0.20±0.04), even after standardizing GEFV predictions to be comparable to GEBV. Results revealed great potential for using GWFS in breeding programs that select families. A significant drop in genotyping costs as one sample per family is needed would allow the application of GS in minor crops.