Directed strategies for improving phenotypic traits

A method for improving at least one phenotypic trait of interest in plant offspring by selecting combinations of at least three individuals within a breeding population of plants, wherein the combinations have for the at least one phenotypic trait of interest a higher Combined Genome-Wide Estimated Breeding Value (CGEBV) in the offspring for the at least one phenotypic trait of interest, as compared to at least 70% of the other combinations of at least three individuals within the breeding population of plants, wherein the method comprises:

• (a) collecting phenotypic data for the at least one phenotypic trait of interest for each individual within a training population of individual plants, wherein the at least one phenotypic trait of interest is a quantitative trait selected from fruit size, fruit count, yield, plant height, relative growth speed, flowering time, germination rate, leaf area, disease resistance, yield components, and biochemical composition;
• (b) collecting genotypic data in the form of genotypic markers for each individual within the training population using a molecular marker technique selected from the group consisting of detection of SNPs, detection of RFLPs, detection of SSR polymorphisms, detection of RAPDs, the detection of indels or CNVs, and detection of AFLP, and attributing to each allele of a plurality of loci of each individual plant, an allele substitution effect for the at least one phenotypic trait of interest;
• (c) providing a genotype/phenotype relationship model for the training population of individual plants, wherein the model estimates for a given genotype of an individual plant what the quantitative contribution is of the allele substitution effects of the plurality of loci on the at least one phenotypic trait of interest;
• (d) genotyping each individual plant within a breeding population in the same way as in step (b);
• (e) calculating for each individual plant within the breeding population the allele substitution effect (P) for each allele of the plurality of loci by using the genotype/phenotype relationship model of step (c), and correcting said allele substitution effect for recombination probabilities with flanking loci (F), wherein a corrected allele substitution effect (S) for a locus is calculated according to formula S=PF, wherein the correction constitutes multiplying said effect with the probability that the allele is transmitted to the offspring in case said allele has a positive allele substitution effect, and multiplying said effect with the probability that the allele is not transmitted to the offspring in case said allele has a negative allele substitution effect;
• (f) determining the CGEBV in the offspring for the at least one phenotypic trait of interest for each combination of at least three individual plants within the breeding population by summing for each combination of at least three individuals for each locus of the plurality of loci in the offspring the highest corrected allele substitution effects calculated in step (e), wherein the CGEBV is determined according to formula
• Σi=1p max(sϑ1,i,sϑ2,i,sϑ3,i);
• (g) selecting the combinations of at least three individual plants within the breeding population that provide for the at least one phenotypic trait of interest CGEBVs in the offspring that are higher than at least 70% of the CGEBVs in the offspring of other combinations of at least three individual plants within the breeding population,
• (h) intercrossing the selected combination of at least three individual plants to produce an offspring population,
• (i) measuring the value of the quantitative phenotypic trait for individual plants in the offspring population, and
• (j) propagating or having propagated offspring exhibiting an improved value of the quantitative phenotypic trait as compared to the at least three individuals.