Heritability of complex disease Flashcards
What can we conclude from the genetics of more complex diseases compared with single gene disorders?
Predisposition to a condition is determined by the complex interplay of the genotype at many different gene, and the effects of the environment, all possibly interacting with eachother in complex manners.
‘Is the disease caused by something in the genes, or something in the environment?’ WHy does this statement not make sense?
Environmental - Speaking english
Genetics - The ability to speak english
What is the ideal question to ask about a disease/trait
How heritable is it…
What does heriatiblity measure?
Measures the extent to which variation in a trait within a particular population is due to variation in genetics.
How do we first go about measuring heritability?
Need to break down variation of the trait/disease/phenotype into genetic and environmental components.
Whats the most common model to explain vairance?
ACE model
A - Additive Genetic variation
C - Common environment
E - Non-common/non-shared environment
What is deviance
Each individual differs from thr population average by some amount#Deviance is related to variance
In individuals, what is their deviance a result of?
Genetic deviance
Environmental deviance
What is genetic deviance
How they genetically differ from the average
What is environmental deviance
How much individuals environmental causes differ from the average.
How do we look at genetic influence on phenotypes in humans?
Monozygotic twin studies
What is broad sense heritability?
refers to all genetic effects – how your DNA contributes to phenotype
What is narrowsense heritability?
refers to just the additive effects.
What is a problem with Monozygous twin studies?
Assumes that the environmental difference between monozygoutic twins is representative of the environmental difference in the population. Which isnt true.
Twins tend to share more of their environment with each other than two random membraer of the population.
What is the shared/common environment?
This extra environment twins have in common. Represented by C in ACE model.
What is the non-shared/non-common environment?
Way in which the environment of twins differ. Represented by E.
How do we disentangle the effects of the shared and non-shared environment?
What can we assume?
Use dizygotic or non-identical twins.
We assume that dizygotic twins share the same amount of environment with each other as monozygotic twims. But they only share the same half as much of their genetics with eachother as monozygotic twins.
What can we conclude from dizygotic twins about genetics and environment when comparing to monozygotic twins?
IF dizygotic twins differ from eachother more than monozygotic twins do, it must be down to this different in genetics.
If dizygotic twins share half their genetics (on average), instead of all their genetics, we might conclude that the total amount of genetic contribution is twince the difference in the similarity of mono- and di- zygotic twins.
Explain dominance relationships
WHere the heterozygote has the same phenotype as one or other of the homozygotes.
Explain epistasis
Where the effects at one locus depend on the effects at another locus.
What are dominance relationships and epistasis cases of?
Where genetics are non-linear, and if either are present, then the twin method will overestimate the additive genetic component.
How can you detect if the genetics are non-linear?
To look at value of c
If 2rDZ - rMZ < 0 = there must be dominance or epistasis effects at work (although C>0 does not necessarily mean there is not).
What do twin studies measure?
How is the estimate inflated?
Measures broad sense heritability (H2), under the assumption that all genetic effects are additive. (H2=h2)
Estimate will be inflated if there are Dominance or epistatic effects.
If 2𝑟_𝐷𝑍−𝑟_𝑀𝑍<0, then 𝐶<0, , what can we assume? What do we then use?
If 2𝑟_𝐷𝑍−𝑟_𝑀𝑍<0, then 𝐶<0, and we therefore assume that there must be dominance effects at work, so we use an ADE model
Can’t estimate C and D at the same time, only one or the other.
Is it true that 𝐶_𝑀𝑍=𝐶_𝐷𝑍?
In utero environment
Are MZ and DZ twins treated the same?
Whats a second assumption about twin methods?
Is that monozygotic and dizygotic twins share the same amount of common environment
Why might MZ and DZ twins not share same amount of common environment?
Parents might trear mono-zygotic twins more alike than DZ twins. MZ twins also have a different in-utero environment to DZ twins.
What are the alternatives to twin studies?
Add
Example of a study estimating geneitc heritability
- 15 million swt of twins - 17,804 traits.
- Looked at : Height, blood pressure, metabolic status, depression and personality disorders.
- They calculated A, C and E using fixed effect linear models
WHat are fixed effect linear models?
In statistics, a fixed effects model is a statistical model in which the model parameters are fixed or non-random quantities.
Fixed effects is a statistical regression model in which the intercept of the regression model is allowed to vary freely across individuals or groups.
What did the study of estimating genetic heritability find?
Found whule some traits have low heritability and some with high heritability - there was a strong average at around 0.5
Basically, on average additivt genetics effects accounts for half of the variation in traits, and many traits were close to this average.
What also did they find about twins?
Found 69% of cases the data was compatibility with a common environment effecr of 0.
Twins shared no more of their relevant environment than any two randon people.
Although did find stronger common environments where expected - like more psychological traits.
Describe DISCOTWIN T2D study
35,000 twins from across Europe.
Fpund heritability of type 2 diabetes (T2D) in adults over 45 was 72% - pretty high!
Washington state twin registry T2D study - what was concluded
Small number of students
Found heritability of 52% in adults over 45,
A lower heritability of 22% in people under that.
From the same study using twins from the same state, find different populations (younger and older) have different heritabilities.
Could be due to chance and small numbers?
Botnia T2D study
Small study from finland.
Found heratibility of 69% in adults aged 35-60, but if looked at participants under 75% they found heritability of 31%
What are possible explanation of the differences in these 3 studies?
Genetics of US and european populations are different
Obesity and food environment have a strong effect on diabetes and that the food environment is different between Europe and Us.
Obesogenic environment more common is US.
Environment conditiond lead to diabetes are less common in europe, therefore chance of getting diabetes is more down to you genes alone.
WHat is the main message
and how do we estimate this?
We cannot assess ‘how genetic’ a disease is, only how much variation n it is due to genetic variation, and is specfic to a particular time and place.
Using the ACE model
What does the ACE model assume?
that Mz and DZ twins share the same amount of environment, and all genetic variation is additive ( i.e there is no dominance or epistasis).
How could a certain amount of heritability be explained by genes and variants?
Genetic architecture.
Genes and/or variants could each have a small effect, or a small number of genes/variants having large effect: what we called the genetic architecture of a condition.
What needs to be thought about when looing for genes invovled with complex disease
Two different measures of how important they are, or what fraction of the heritability is down to that variant - how common is an allele and how big is the effect of having that allele.
Different measures of how big the effect of an allele is
Genotype Relative RIsk
The Odds Ratio
Genotype Relative RIsk
What is the probability of having the disease if you have the risk genome type relative to the probability uif you have the pretective genotype.
.. so if you are twice as likely to get the disease if you have A rather than T, the GRR is 2.
The Odds Ratio
The odds of getting the disease if you have the risk genotype relative to the odds if you don’t
Here the odds is the number of people with the disease divided by the number of people without it.
What are the Genotype relative risk and Odds ratio talking about?
What else needs to be considered?
Disease vs non-disease
Continuous traits like height, BMI, blood-pressure - in these cases the effect size is measured by Beta.
What are the Genotype relative risk and Odds ratio talking about?
What else needs to be considered?
Disease vs non-disease
Continuous traits like height, BMI, blood-pressure - in these cases the effect size is measured by Beta.
What is Beta?
Give an example
The amount of trait changes for each additional allele.
Example - for height, a Beta of 1 might mean height increases by 1cm for each allele
What does an allele that has a big effect have?
An allele that has a big effect has a big genotype relative risk or big odds ration - but that doesnt mean that it explains a lot of the heritability.
How can different possibilities for disease allele be visualised?
On a plot with Allele Frequency on one axis and effect size on the other.
When does a disease behave as a mendelian disorder?
If effect size is big, such that people with the risk allele almost always have the disease and those without it don’t
What is ideal to identify?
Give 3 examples
Identify genes that are both common and have a large effect. If such genes exist, they have been identified.
DQB1 in type 1 diabetes
ApoE e3 allele in Alzheimers
Mutation in CFH gene
Describe DQB1 in type 1 diabetes
A change of Asp57 to any toehr amino acid has a very large effect.
Homozygous Odds ratio is 18, but 3% of the population is homozygous, only 0.3% of population get T1D.
Genetics is not predestination.
This is the largest effect size found of any allele for a complex disease, but only 10% of those who are homozygous for it will get the disease.
Describe ApoE e3 in Alzheimers
Has an heterozygous Odds ratio of 2.5
Has an homozygous Odds ration of 13.
Allele frequency of 16%
CFH gene, Wet Mucular degeneration
CFH gene offers protection from Wet macular degeneration.
Adds ratio of 5 - those with the risk allele have odds 5 times more than those without.
These effects were discovered in the pre-GWAS era.
WHat are the group of alleles addressed by Genome-wide association studies.
Disease alleles are common, but only have a small effect.
