Allelic diversity for candidate genes and association studies: methods and results

The increasing ease with which molecular markers can be generated makes it possible for plant geneticists to use these genomic technologies for better exploitation of the available genetic variation in breeding populations. Identifying markers based on conventional bi-parental mapping populations is most likely not the best way to implement a marker assisted selection (MAS) program, although this approach is useful for introgression of alleles from wild germplasm. Instead, association mapping may be used in a more practical approach, by measuring both phenotypes and markers directly on the plants in the breeding nursery. Conventional quantitative trait loci (QTL) mapping enables one to identify chromosomal regions of 5–20 cM containing genes underlying the trait of interest. However, that still leaves several hundred potential candidate genes. Association mapping enables the exploitation of the wider genetic diversity and incorporate a larger number of recombinations. Synthetic populations used for genetic improvement of self-incompatible crops including many forage and turf species, present a useful tool for incorporating association mapping and genotype building using molecular markers. This is particularly true for traits that have not previously been selected for, since linkage disequilibrium (LD) is less likely to have been built up. We show some preliminary data from a experiment to illustrate population structure, LD and associations with candidate genes in synthetic populations not previously selected for this trait. Some recent research on association analysis in perennial ryegrass and clovers are also reviewed. We also briefly describe genomic selection (GS) that can predict the breeding values of lines in a population by analyzing phenotypes and high-density marker scores as a way to incorporate MAS into the breeding process.