Tipsters: I was on the right track on my last post, but, I may have said some
things wrong. Let me try again.
"Additive" refers to things like height (and maybe IQ) where the more genes
for height you have, the taller you are. "Nonadditive" refers to cases like
Mendelian recessive-dominant at a single locus. It also includes interactions
between gene sights. Plomin gives the example of "beauty." Rather than being
the result of alot of "beauty genes", beauty is the result of interactions
between separate parts. A long nose, for instance, may be great with one mouth
but lousy with another. These interactions cannot affect MZ twins because MZs
are identical for additive and nonadditive effects. DZ twins, in contrast, are
made _different_ by non-addtitive effects. Stephen's example of a single
recessive-dominant fits in here as an extreme example, where the interaction
creates variability in offspring (the point: genetics not only makes us
similiar - it mnakes us different within families). So, while nonadditive
effects do not affect r(mz) they make r(dz) _lower_. Because they made r(dz)
disproportionally low, 2*(r(mz)-r(dz)) overestimates H. Plomin claims the
r(mz)-r(dz) difference contains half the additive effect, but nearly all the
nonadditive effects. Plomin calls the twin method an estimate of _broad_
heritability because it includes additive and nonadditive effects. In
contrast, H obtained from adotption studies _underestimates_ genetic effects,
and estimates _narrow_ heritability because it only contains additive effects.
This is because all first-order relative relationships such as mother-child or
sibling-sibling, are made lower in the face of non-additive effects. Doubling
the biological mother-adopted child r underestimates the genetic effect. Whew!
I'm not sure this gets to the very heart of Stephen's question - or just
rephrases it. At least, it's a reminder that the twin and adopion method
estimates are effected in different ways by non-additive genetic effects, and
that a simple recessive-dominant situation is an example of a "nonadditive"
genetic effect.
Source: Plomin, R. (1990) Nature and Nurture: An Introduction to Human
Behavioral Genetics. Brooks/Cole.
--
* John W. Kulig, Department of Psychology ************************
* Plymouth State College Plymouth NH 03264 *
* [EMAIL PROTECTED] http://oz.plymouth.edu/~kulig *
*******************************************************************
* "I was made in secret and curiously wrought in the lowest *
* parts of the earth" - Psalm 139 *
*******************************************************************
