Agriculture Reference
In-Depth Information
controlled crosses or transgenic expression important for complement-
ing AM studies.
There is little advantage of association over linkage mapping in
populations for which LD is extensive and LD decay does not occur
rapidly across most of the genome, which appears to be common in
many self-pollinating species. In this case no reliable correlations can be
found between speci
c genes and traits, rather a whole region of the
genome is associated due to the lack of haplotype block breakup. Most
crops have a degree of LD due to domestications genes or adaptation
suites inherited from genetic bottlenecks. These genes can also lead to
spurious associations. This problem is particularly apparent when trying
to map traits that have been subject to local adaptation like
flowering
time because variation in these phenotypes is highly correlated with
allele frequency differences between populations (Myles et al. 2009).
Disease resistance loci consisting of tandem repeats of similar leucine-
rich repeat genes might have too rapid a decay in LD and unmanageable
insertion
-
deletion events in their region, which may also make these
traits dif
cult to evaluate with AM approaches. However, quantitative
resistance based on underlying genes for disease resistance rather than
fast-evolving sensor-decoy resistance (R) genes may be obtained with
AM analyses.
B. Alternatives
In the case of crops where there are constraints to AM or low resolution
caused by high LD, there are some alternatives that still take advantage of
AM, but solve the problems of population structure and rare alleles.
These might be alternatives that empower AM in most crops as well.
They rely on identifying good combinations of alleles found usually in
breeding lines and crossing these into a multiparent, structured popula-
tion that can be used to detect marker
trait associations more effec-
tively (Kover et al. 2009). Here we present two speci
×
c alternatives that
we have mentioned earlier inmore detail: NAMpopulations andMAGIC
populations.
NAM populations were
first developed in maize based on 25 related
families resulting from test crosses with the same inbred and each family
with 200 RILs developed to an advanced generation (McMullen et al.
2009). The common parent was a well-characterized genotype and the
other 25 parents were diverse genotypes contributing a wide range of
alleles. This strati
cation allows linkage mapping with substantial
statistical power and still captures genetic diversity but requires
the phenotypic screening of large numbers of lines (5000 in total!). A
