In principle:
Extend single-locus multi-locus quantitative models
1AA
: 2Aa : 1 aa (1AnAn : 2Anan : 1anan)n normal distribution
Genotype
/ Phenotype correlation Heritability
Genotypic expression depends on environment
Heritability
(h2) estimates proportion of phenotypic variance due to
genetic variance
"Is It Genetic?" :
Genotype / Environmental interaction is variable &
unpredictable
Quantitative
Trait Loci (QTL):
genetic variation can be correlated with phenotypic
variation
Variation can be quantified
(review)
mean standard
deviation:
variance: 2
Variation follows "normal
distribution" (bell-curve) iff
Multiple loci are
involved (quantitative)
Each locus has about equal
effect (additive)
Each locus acts independently
(interaction is minimal)
Ex.: Suppose a trait is influenced by 5 loci, each with two
alleles A & a
35 = 243 genotype classes in 11 phenotype classes
variation continuous
Phenotypic variance ( 2P ) has two sources:
genetic (2G) & environmental (2E) variance
Variance is additive:
2A+B
= 2A
+ 2B
phenotypic variance 2P = 2G + 2E + 2GxE
where 2GxE is interaction variance, if 2G and 2E are not independent
If 2G and 2E
independent (or assumed so), 2GxE ~ 0
additive variance 2A = 2G + 2E
Heritability h2 = 2G / 2A = 2G / (2G + 2E)
"heritability
in the narrow sense"
genetic component of the additive variance
heritability (h2) is the fraction of the (additive)
phenotypic variance due to genotypic
variance
ignoring
interaction
variance 2GxE
genotype / phenotype
relationship differs in
different environments.
Ex.: same strain of corn
produces different yields in different fields
Artificial breeding indicates that phenotypic variation is
(often) highly heritable
Artificial selection
on agricultural species
Commercially
useful
traits
can
be
improved
by
selective breeding
Common
Garden experiments
Correlation shows association between variables (cf. regression analysis)
Mid-parent
value
Offspring / mid-parent correlation
estimates heritability
For many quantitative traits in many organisms: h2 = 0.5 ~ 0.9
Heritability
is typically calculated in a single
environment.
The Norm of Reaction mediates genotype through environment to produce phenotype
Variation within groups: Is variation 'genetic'
?
Variation among groups: Are differences 'genetic'?
Genetics, Heritability, & Society
"Is it Genetic?"
Common Myths about Genetics
Myth 1: Genetic diseases are determined
by the presence or absence of a "gene
for a trait"
Everyone has the same set of genes: persons with
alternative forms a "genetic trait" have different forms
of the gene (alleles)
Ex.: Allelic variants of the PAH gene
result in a non-functional enzyme unable to
metabolize phenylalanine (Phe)
If not treated by
reduction of dietary Phe, this results in the disease
Phenylketonuria (PKU)
PAH is not a gene "for" PKU ,
rather a gene for an enzyme that prevents PKU
Ex.:
The typical function of "cancer genes" is
to prevent cancer: some allelic variants fail to
perform this function properly
"Removal" of
these genes would not "cure cancer": removal of
defective spark plugs does not allow a car to start
Myth
2: Genetic traits always follow the Mendelian Model
Ex.: If rrww peas are always green
and wrinkled
Then persons with certain
alleles "for" BRCA1
with always get breast cancer
Many traits follow a Quantitative Model, with 10s ~ 100s
of genes each contributing a small effect
Ex.: Parental x Offspring
differences in height
Ex.: Predisposition to obesity
within families.
Myth 3:
Heritability is a measure of Mendelian inheritance
Ex.: Traits with high "heritability" need not be genetically
pre-determined
*** Heritability (h2)
is a specific mathematical concept :
the fraction of total
measurable variation (variance) due to genetic
variance
h2
2Genotype
/ 2Phenotype
not
including 2E
& 2GxE
the assumption that h2 =
2P
= 2G ignores 2E
& 2GxE
Myth
5: That which is "genetic" is fixed &
unalterable
Ex.: Phenylketonuria is
perfectly "genetic" in the Mendelian manner AND perfectly
treatable
Ex.:Heritability, IQ, &
Education
IQ test scores in Homo: h2 =
0.7~0.8 within groups
Highly
heritable traits can be
highly modifiable by change of
environment if 2GxE is
large
All text material ©2016 by Steven M. Carr