Agriculture Reference
In-Depth Information
Somers et al., 2007). Given a wheat genome map
length of 2,500 cM (Somers et al., 2004), then
2,500 markers (1 marker cM
−1
) would be required
to exceed the level of LD decay. This is the
primary driver behind SNP development, as it
could provide the needed marker density for effi -
cient association mapping, although the concerns
with low polymorphism levels with SNPs are still
untested.
The advantage of association mapping versus
QTL analysis in biparental crosses lies largely in
population development. Populations for associa-
tion mapping are simply a collection of existing
wheat germplasm that can be characterized by
genotyping. The collection makeup is easily con-
trolled both for size and genetic content. The
population can be diverse, such as a global collec-
tion of wheat accessions (Maccaferri et al., 2004),
or it can be restricted to accessions adapted to
a certain country or region (Chao et al., 2007;
Somers et al., 2007). Since the population under-
goes genotyping, it can be tailored to contain a
diverse collection of alleles across the genome or
have limited variation at certain loci. In contrast,
biparental crosses and derived populations require
time (1-4 years) and effort to produce; selection
of the two parents determines the makeup of the
mapping population. The QTL analysis is then
restricted to the pairs of alleles segregating at each
locus in the population. Hence another advantage
of association mapping is the ability to analyze
associations of traits with multiple alleles, includ-
ing rare alleles.
The fi nal advantage of association mapping
populations concerns the number of meiotic
events represented in the collection of wheat. In
most cases, the accessions within an association
mapping population will each have been derived
through many meiotic events. Thus a high-reso-
lution genotypic comparison of wheat accessions
within the association mapping population will
show multiple crossover events within a defi ned
marker interval. In contrast, a typical doubled-
haploid mapping population derived from a bipa-
rental cross will have undergone a single meiotic
event, and the number of crossovers is limited. As
a result, if a high-density marker system is avail-
able for both populations, the association mapping
population has more potential to provide precise
marker-trait associations with fewer individual
lines in the population, compared with biparental
crosses.
Gene expression analysis
One of the newest applications of QTL analysis
is mapping gene expression patterns. Microarray-
based analysis of gene expression, such as the
Affymetrix platform and the Affymetrix wheat
gene chip (http://www.affymetrix.com),
has facilitated highly parallel quantifi cation of
individual genes under controlled biological
experiments. The Affymetrix wheat gene chip has
in excess of 55,000 genes represented. Expression
of individual genes can be regarded as quantitative
phenotypic data on this platform. This is referred
to as
genetical genomics
(Jansen and Nap 2001).
Thus if gene expression data is collected from a
biparental mapping population, then an SSR map
of this population can be used to perform QTL
analysis on individual gene expression patterns.
Further, if the mapping population is character-
ized for agronomically important traits or seed
quality attributes, the gene expression QTL
(eQTL) can also be aligned with these traits.
This approach was taken recently with a
doubled haploid population that was fully mapped
with SSRs and fully phenotyped for many agro-
nomic and seed quality traits (McCartney et al.,
2005a, 2006). The microarray gene expression
analysis of this same population used mRNA
extracted from developing seeds (5 days postan-
thesis). The result was mapping >500 eQTLs
across the genome. The eQTL results, when
aligned with existing phenotypic traits, should
succeed at identifying areas of the genome and
candidate genes that control seed quality traits
(Jordan et al., 2007).
FUTURE PERSPECTIVES
More than 20 years has passed since the develop-
ment of the fi rst, crude molecular genetic maps of
wheat, beginning with RFLP markers and advanc-
ing to SNP and DArT markers, with increasing
