Chapter 7 - Estimating Genetic and Environmental Influences

Heritability

The amount of phenotypic variance accounted for by genetic differences among individuals (population)

Additive Traits

Genes whose effects simply add up to build a trait

Dominance

When one gene masks or overpowers the effects of another gene

Gene interaction

When multiple different genes interact with each other to create complex effects

Broad-Sense Heritability

Measures how much of the differences we see in a specific traits across a population is caused by every possible genetic factor combined (i.e., Dominance, Gene Interaction, and Additive Traits)

What does it mean when heritability equals 100%?

All variation could be explained by genetic differences between individuals

Effect Size

Measures how much of the differences we see between people across an entire population is caused by their genes. It answers the question of how much genetics actually contributes to a trait

Environmental effects are often __, but genetic effects are rather __

small; large

h²

Genetics. They focus strictly on additive genetic effects, which are the genes passed down directly from parents. The symbol h² represents the proportion of variance caused by these additive genes.

c²

Environment. They divide the environment into "shared" and "non-shared." The symbol c (or c²) stands for the shared environment—the things siblings experience together that make them similar, like growing up in the same house with the same parents

Equation for DZ twins

r = h²/2 + c²

Equation for MZ twins

r = h² + c²

Falconers Formula

h² = 2(r_mz - r_dz)

Narrow-Sense Heritability

Because identical twins share twice as much genetic material as fraternal twins, you can calculate the total narrow-sense heritability of a trait simply by taking the difference in their similarities and multiplying it by two

Concordance

Index of risk. Used to estimate either/or disorders (i.e., you either have it or you don’t). Simply measures the probability that if one twin has a disorder, the other twin will also have it.

What does it mean if the concordance rate is the exact same for both identical twins and fraternal twins?

Zero Heritability. It means that sharing 100% of your genes does not increase your risk of getting the disorder compared to sharing only 50%. Therefore, the genetic influence is non-existent, and heritability is zero

Liability Threshold Models

This model assumes there is a hidden, continuous spectrum of genetic risk (liability) in the population, and a person only develops the disorder once their risk crosses a certain "threshold". This model mathematically translates the concordance percentages into standard correlation scores

Structural Equation Models (SEM)

Advanced statistical models used to find the most mathematically accurate "best fit" to estimate the true effect size of genes and environment on behavior

SNP Heritabiltiy

A modern method that bypasses family comparisons entirely. Instead, it estimates heritability by looking directly at the tiny variations in the DNA code (Single Nucleotide Polymorphisms, or SNPs) between massive groups of completely unrelated individuals

Peter Visscher (2008)

Reviewed methods of estimating heritability

Applications of Heritability (Peter Visscher Study)

• Making Comparisons: Helps researchers understand how much of the variation within a specific population is driven by genetic differences

• Predicting Breeding Success: Predicts how well natural or artificial selection will work.

• Estimating Disease Risk: In medicine and research, it is used to estimate the genetic risk of a person developing certain diseases

Misconceptions of Heritability (Peter Visscher Study)

• It’s the exact proportion of a trait passed down: Heritability does not mean you inherit exactly a certain percentage of a trait from your parents

• It means you genes are you destiny: It’s not strict genetic determinism. A person’s observable trait is a good predictor of their genetic makeup

• Explains differences between different groups” It only measures the differences within one specific population

• High heritability means there a re massive major genes: It’s a myth that a highly heritable trait must be controlled by a few genes with huge effects

• It tells us what is possible: Heritability only describes “what is” meaning the current genetic differences in the current environment, not what “could be” if conditions were to change

Galton’s Parent-Offspring Regression

A straightforward mathematical way to measure how much of a trait is passed directly from parents to their children. Calculate the “mid-parent” score (the average of the mother and father’s trait) and compare it to the “mid-offspring” score (the average trait score of their children). After points are plotted, a regression line is drawn. The slope equals the narrow-sense heritability (h2). The slope is a direct estimate of the additive genetic variance.

What does a steep slope mean?

The children’s traits closely match the parent’s traits, meaning heritability is high

Estimating Heritability from Inbred Strains

Estimating Heritability from Inbred Strains This method is used in lab experiments or agriculture, where researchers can control the breeding process.

• Genetic Uniformity: By continually inbreeding animals or plants, researchers create a strain that is entirely genetically identical.

• Environmental Variation (Ve​ or VF1​): If you look at a highly inbred strain, or their first generation of offspring (F1), they all share the exact same genes. Therefore, any physical differences we see between them must be completely caused by the environment.

• Total Variation (Vp​ or VF2​): In the second generation of offspring (F2), genes begin to segregate and mix up again, which creates a combination of both genetic and environmental differences. This gives researchers the total variation.

• The Formula: To find heritability, researchers use the formula h2=(VF2​−VF1​)/VF2​. In simple terms, they take the total variation of the mixed group (VF2​), subtract the purely environmental variation of the genetically identical group (VF1​), and then divide the result by the total variation. What is left over is the exact proportion of variation caused by genetics.

A method used in lab experiments, where researchers can control the breeding process.

• Genetic Uniformity—Continually inbreeding animals leads to a strain that is entirely genetically identical (F1 generation shares same genes or VF1), so any physical differences are completely caused by the environment

• Total Variation—In the second generation (F2 or VF2), genes begin to segregate and mix up again, which creates a combination of both genetic and environmental differences = Total Variation

• Formula—(VF2 - VF1) / VF2 = Exact proportion of variation caused by genetics

Breeder’s Equation

Used to estimate heritability through a “truncated” breeding experiment. Selectively breeding the highest-scoring individuals to see how much of a specific trait is actually passed on to the next generation

• Step 1—Stating Average (X0): First, measure the average score of a specific trait across entire starting population

• Step 2—Cutoff and Selection (X1 and S): Next, create a strict cutoff point and only allow the top performers above the line to breed. Calculate the average score of just chosen (X1). The difference between the chosen parents and the original population is called the Selection Differential (S).

• Step 2—Offspring Results: (X2 and R): Finally, measure the average score of the second generation (offspring of elite breeders). The difference between these new offspring and the original population is called Response to Selection (R)

Equation: h² = R / S = heritability of the trait

Individuals vs Population

For any single person, both their genes and their environment are absolutely indispensable—you can’t build a person without both. So heritability doesn’t measure the phenotype of a single person, instead the genetic contribution to the differences between people within a larger group.

Height Example

If you hear that the heritability of heigh is 0.9, this doesn’t mean a specific person grew to 90% of their total heigh because of their DNA, it means that 90% of the height differences observed among all the people in that population are caused by genetic differences

Old Assumption vs New Evidence for Parent/Offspring Resemblance

The old assumption is that children act like their parents mainly because of the way they were raised. However, modern twin and adoption studies have challenged this idea, revealing that a large portion of what we previously thought was just “family resemblance” caused by the environment is actually driven by shared genes.

Shared Environment

The non-genetic influences that makes family members similar (e.g., experience, neighborhood, parental education, parenting styles

Non-shared Environments

Refers to all non-genetic influences that are independent (uncoorelated) for family members

Variance Partitioning

P = A + D + C + E;  where (P) phenotype is a function of (A) additive genetic effects, (D) dominance genetic effects,  (C) shared environmental effects, and (E) non-shared environmental effects

• Specifically, C is the estimate of family resemblance not due to genetics and E is the remain variance not explained by the other parameters

NEAD Study

A study for identifying specific non-shared environmental effects (real-world factors that cause differences in siblings). They interviewed 720 families with two siblings between 10-18yrs old. Results show sibling correlation were modest, parent-to-child and child-to-parent correlations were also modest. Parent correlations were high suggesting less value for assessing non-shared environmentality

Single Nucleotide Polymorphisms (SNPs)

The most common type of human genetic variation, occurring when a single DNA building block (nucleotide) is changed. They act as crucial biological markers throughout the genome to help detect individual genotypes, locate disease-associated genes and estimate heritability.

GWAS

Combines quantitative and molecular methods to scan the genomes of many individuals to detect specific genetic variations associated with diseases. It highlights the “missing heritability problem”, which occurs when these identified DNA variations fail to fully explain the known genetic variance of complex traits like schizophrenia

Two Methods of SNPs

• GREML: Uses a mixed regression model of fixed and random effects to estimate heritability directly from the genotypes of unrelated individuals

• LDSC: Estimates the heritability using a regression slope of linkage diequilibrium scores, meaning it does not even require the original GWAS genotypes

SNP vs Quantitative Methods in Twin/Adoption Studies

Comparison of SNP to quantitative methods show that SNP estimates of heritability are smaller than ones obtained in twin/adoption studies

Multivariate Analysis

Investigates the genetic and environmental influences on multiple traits simultaneously. This helps researchers estimate whether the exact same underlying genetic or environmental factors are driving completely different traits