146
E. A. Thompson
6.4
IBD-Based Genetic Mapping
Any genetic mapping procedure aims to detect the genomic locations of DNA
variation underlying a trait of interest, by reference to a genetic map of markers
that have known locations in the genome. An association test directly considers
the dependence between marker genotypes (X) and trait phenotypes (Y). However,
these allelic associations arise from the descent of DNA from common ancestors to
different individuals within a population or to different populations. It is therefore
useful to consider the associations between X and Y through the lens of descent Z
and specifically patterns of IBD among individuals observed for the trait inferred
at locations across the genome. Throughout this section we assume that Z contains
all the information needed for analysis of association between X and Y: that is, we
assume that X and Y are conditionally independent given Z.
6.4.1 Mapping from IBD in Pedigrees
We first consider the pedigree context, in which prior probabilities of IBD are
provided by the specified pedigree relationships among individuals. At locations
across a chromosome, the genetic marker data X provide probabilities of IBD or
realizations of the inheritance vectors which determine the location-specific descent
of DNA through a pedigree (Sect. 6.3.2).
In genetic mapping, the use of location-specific IBD in affected individuals of
known pedigree relationship has long been established as a powerful approach. This
may be IBD in affected sib pairs (Suarez et al., 1978) or more general relative pairs
(Weeks and Lange, 1988) or even the two parental gametes of individuals affected
by a rare recessive trait (Lander and Botstein, 1987). In each case, individuals
sharing the trait have increased probability of sharing genome IBD at causal loci.
Observation of the same genome regions showing IBD across multiple pairs of
affected relatives provides the linkage signal. Significance of the observations can be
readily assessed, since the known pedigree relationship provides the null distribution
of any IBD-based test statistic.
IBD-based tests for linkage have been extended to larger groups of relatives,
and a wide variety of test statistics have been developed (McPeek, 1999). Joint
sharing of genome IBD by multiple affected relatives generally provides stronger
evidence, since the pedigree-based prior probability of this multiple sharing event
is generally smaller. An advantage of an IBD-based approach is that it is relatively
robust to allelic heterogeneity: within a pedigree the affected individuals are likely
to carry the same causal mutation IBD. If there are large pedigrees with multiple
affected individuals available, even locus heterogeneity is a lesser concern, since
even a single pedigree can provide sufficient evidence of linkage.
In the genetic mapping of loci underlying quantitative traits (QTL), IBD-based
approaches also have a long history. Haseman and Elston (1972) developed an
E. A. Thompson
6.4
IBD-Based Genetic Mapping
Any genetic mapping procedure aims to detect the genomic locations of DNA
variation underlying a trait of interest, by reference to a genetic map of markers
that have known locations in the genome. An association test directly considers
the dependence between marker genotypes (X) and trait phenotypes (Y). However,
these allelic associations arise from the descent of DNA from common ancestors to
different individuals within a population or to different populations. It is therefore
useful to consider the associations between X and Y through the lens of descent Z
and specifically patterns of IBD among individuals observed for the trait inferred
at locations across the genome. Throughout this section we assume that Z contains
all the information needed for analysis of association between X and Y: that is, we
assume that X and Y are conditionally independent given Z.
6.4.1 Mapping from IBD in Pedigrees
We first consider the pedigree context, in which prior probabilities of IBD are
provided by the specified pedigree relationships among individuals. At locations
across a chromosome, the genetic marker data X provide probabilities of IBD or
realizations of the inheritance vectors which determine the location-specific descent
of DNA through a pedigree (Sect. 6.3.2).
In genetic mapping, the use of location-specific IBD in affected individuals of
known pedigree relationship has long been established as a powerful approach. This
may be IBD in affected sib pairs (Suarez et al., 1978) or more general relative pairs
(Weeks and Lange, 1988) or even the two parental gametes of individuals affected
by a rare recessive trait (Lander and Botstein, 1987). In each case, individuals
sharing the trait have increased probability of sharing genome IBD at causal loci.
Observation of the same genome regions showing IBD across multiple pairs of
affected relatives provides the linkage signal. Significance of the observations can be
readily assessed, since the known pedigree relationship provides the null distribution
of any IBD-based test statistic.
IBD-based tests for linkage have been extended to larger groups of relatives,
and a wide variety of test statistics have been developed (McPeek, 1999). Joint
sharing of genome IBD by multiple affected relatives generally provides stronger
evidence, since the pedigree-based prior probability of this multiple sharing event
is generally smaller. An advantage of an IBD-based approach is that it is relatively
robust to allelic heterogeneity: within a pedigree the affected individuals are likely
to carry the same causal mutation IBD. If there are large pedigrees with multiple
affected individuals available, even locus heterogeneity is a lesser concern, since
even a single pedigree can provide sufficient evidence of linkage.
In the genetic mapping of loci underlying quantitative traits (QTL), IBD-based
approaches also have a long history. Haseman and Elston (1972) developed an