A Breeding Plan for Molecular Markers.

Genotyping is no longer the bottleneck for genetic improvement. Therefore, gene information might not depend on indirect measurements via molecular markers. Breeders focus on complex phenotypes while molecular biologists focus on simple genotypes. The lack of interaction between breeders and molecular biologists has caused a significant funding imbalance with severe consequences on the application of basic science and the training of the next generation of breeders. A breeding plan for molecular markers is proposed to select top parents (e.g. with all loci in a desired state) carrying up to 30 loci with two alleles each assuming genes are known and can be genotyped. With L segregating loci we would generate 2L individuals to genotype for each locus. The basic plan would be:
1)      Select two individuals for crossing that collectively have at least one favorable allele at each locus.
2)      From this subset select the cross that has the highest average frequency of favorable alleles over all loci.
Simulation of this plan would suggest that 2L individuals would be enough to have a high probability of observing at least one pair of individuals that collective have all loci represented in a favorable state. The number of cycles required to select the individual homozygous at all loci would increase with an increase in the number of loci segregating. We consider this plan an essential tool to bring genomics and breeding students together by allowing them to select the best parents for a breeding program knowing that choice of germplasm sources is ultimately the most important decision in a breeding program. Without the right germplasm the technology would not help.