Behavior Genes
603
means that it is possible to construct an
SNP map across regions of interest (e.g. a
region implicated by linkage studies) and
narrow it down using case control stud-
ies to detect LD. Databases can then be
searched for genes within the narrowed re-
gion of interest and SNPs selected within
such genes to explore them further in case
control comparisons. This process, which
in effect is a convergence of the positional
cloning and candidate gene approaches, is
now referred to as the
positional candidate
approach
and has been successful, as we
will discuss below, in implicating several
genes in the etiology of schizophrenia.
4
Applications
Quantitative approaches have now been
used extensively in a wide range of behav-
ioral traits and disorders and increasingly,
molecular genetic approaches are being
applied. Here, we have selected three ma-
jor examples, one a normal trait and the
other two, disorders (or rather groups of
disorders). These are general cognitive
ability or intelligence, mood disorders and
schizophrenia.
4.1
General Cognitive Ability
Intelligence in its various guises is one
of the most extensively studied aspects
of human behavior. There is now a
broad consensus among psychologists that
although various types of intelligence
can be described and tests devised to
measure them, performance on such tests
is strongly correlated with a common
factor called
general cognitive ability
or ‘‘g.’’
Thus, it is has been shown that tasks
such as copying a design with wooden
blocks (visuo–spatial ability), vocabulary
tests (verbal ability), remembering Fgures
(memory), and comparing lists of objects
(processing speed) all correlate with ‘‘g.’’
Commonly used tests provide measures of
general intelligence quotient (IQ) as well
as speciFc abilities including the Wechsler
Adult Intelligence Scale (WAIS) and the
Wechsler Intelligence Scale for Children
(WISC).
Quantitative genetics and cognitive ability
±amily studies of IQ and ‘‘g’’ demonstrate
moderate to high correlations of around
0.4 to 0.5 for Frst-degree relatives. More
distant relatives such as cousins typically
show correlations in the order of 0.15.
Although such a pattern suggests the
involvement of genes, it is not on its
own sufFcient proof. In fact, nearly all of
the twin studies performed in childhood
suggest that the familial correlation in IQ
is partly the result of genes and partly
the result of shared environment. ±or
example, identical twins reared together
show very high correlations for g, on
average around 0.85. This is somewhat
lower, at around 0.7, in identical twins
reared apart. Similarly, DZ twins reared
together show an average correlation for g
of about 0.6, but this falls to an average
0.3 in DZ twins reared apart. It can
be shown using the sort of structural
equation models that have been discussed
earlier that this pattern of correlations
corresponds to a heritability of around
50% with about 25% of the variance
explained by shared environment and
the remainder explained by nonshared
environment.
Adoption studies in childhood are al-
most completely consistent with the twin
results in that adoptive siblings show a cor-
relation for g of around 0.25 compared with
a correlation of 0.5 in biological siblings
previous page 603 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 605 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off