Agriculture Reference
In-Depth Information
confounding effect on breeding and genetic study results, the recent
advances inmarker reliability through effective SNP platforms have seen
continued advancement in this area. Population structure and morpho-
logical traits can be correlated in which case it is unlikely that associa-
tion can be found with a single locus, but it appears that genome-wide
approaches, such as GS, can improve on the prediction associated with
multiple loci and epigenetic interactions (Jannink et al. 2010). Overall,
AM approaches are successful in the majority of cases where phenotypic
variability is found within subpopulations and where candidate genes
are known or marker density is adequate. The analysis of population
structure has been a signi
cant breakthrough for association studies in
plants.
The increase in marker numbers for all crops has allowed association
genetics to overcome the problem of low-marker density in GWA
studies, which can reduce the chances of discovering new loci involved
in a particular trait. Very large numbers of markers may not always be
needed for AM studies depending on the level of LD and the selection of
the types of markers based on their distribution and representativeness
of gene space. For example, if expressed genes are used in GWA analysis,
some savings can be obtained in the number of markers. However, as we
move toward the
”
effective whole-genome genotyping threshold (Deschamps and
Campbell 2010), we see not both GWA, but only the GS approaches
taking off in the public sector as they already have in the private sector,
where funding has been less limiting. Robust statistical corrections for
the effects of population substructure and kinship have been developed
and high LD should permit successful GWA scans using relatively mod-
erate marker densities in some cases. Knowledge about recombinational
history in breeding populations will allow GW methods and AM in
various population types to be more effective.
Another issue of importance is that phenotyping for AM or GS
programs is challenging, because it needs to capture the true phenotype
and all the variants present in a diverse collection of material or breeding
program. This requires good replication and multisite testing of germ-
plasm, which remains expensive. Likewise, phenotype dissection is
required to enhance association power, with phenotyping now being
the principal challenge for AM or GS studies. Genotyping gaps are
solvable and statistical analysis has improved, but in many cases phe-
notyping technologies are still being developed or being rediscovered.
The heritability of a phenotypic trait is likely to affect the discovery and
potential to use genes important to plant breeders, as it always has. The
constraints of heritability affect
“
”
“
1000 dollar
genome or even better the
1 dollar
the theoretical potential
for gene
