“Four Laws of Behavioral Genetics… Behavior and Genomic Medicine… Behavioral Genetics, the Law and Personal Responsibility, Part I… Behavioral Genetics, the Law and Personal Responsibility, Part II – UPDATED… Interview with Irv Gottesman… Acknowledgements, Thank You, and Goodbye…
Supplemental – Genetic Prediction… “
Module C: Behavioral Genetics, the Law and Personal Responsibility, Part I
Module D: Behavioral Genetics, the Law and Personal Responsibility, Part II - UPDATED
Interview with Irv Gottesman
Module F: Acknowledgements, Thank You, and Goodbye
Module G: Supplemental - Genetic Prediction
Module A: Four Laws of Behavioral Genetics
Summarize the whole course through the Four Laws of Behavioral Genetics.
All human traits are heritable;
Shared family environment has minimal impact on individual behavioral differences;
Non-shared environment has major influence on individual behavioral differences;
Human behavioral traits are polygenic.
All human traits are heritable:
For many traits, heritable influence increases as one gets older.
Heritability is to fixed – environments can increase or reduce genetic influences (GxE) on traits.
Shared family environment has minimal impact on individual behavioral differences:
Based on twin and adopted sibling studies.
Except for some traits (e.g. general cognitive ability) but effects decline as one gets older (especially in the adolescence to adult transition).
Suggests shared family environment does not share offspring in the same way but does not mean families have no effect (e.g. large effect but different for different children).
Non-shared environment has major influence on individual behavioral differences:
Studies showing lack of perfect behavioral similarity (correlation and concordance) between MZ twins;
Studies of this also include 10-15% measurement error;
Do not know for sure yet what specific factors in the non-shared environment influences the differences in traits (“gloomy prospect” for psychologists).
Human behavioral traits are polygenic:
Candidate-gene studies and Genome Wide Association Studies (GWAS);
GWAS: some ask whether it is worthwhile to do it if the effects are so small;
Others feel it is as the tracker record is better, and do provide some insights even if the effects are small e.g. FTO gene for obesity has only 1.2 kg (2.6 lb) effect i.e. 1% variance but it provides insights on a different to treat issue in developed countries and that its effects starts early when one is young (i.e. suggests early intervention needed).
Next lesson: behavior and genomic medicine and their interactions.
individuals with elevated risks (e.g. having 20 or 21 of the 22 or 25 genetic variants for schizophrenia);
there will soon be a lot of genetic data available (rapidly declining costs of sequencing a genome).
Examples of treatments according to genotypes e.g. pharmacogenomics has about 100 examples; might this be true in time to come for psychosocial interventions too?
Genetic counselling – lessons from Huntington Disease:
People might not know tests are available, or they might know but not take the tests;
Do we understand the psychological consequences of a positive test result e.g. how it changes over time.
This is an area where psychologists are needed.
Concluding thoughts on genomic medicine: we assume that when someone knows his or her risks, they will take the necessary measures, but we know from psychology that people do not always behave that way – this has implications on whether it will really be preventive.
Module D: Behavioral Genetics, the Law and Personal Responsibility, Part II – UPDATED
Description of the Bradley Waldroup case where the MAO gene variant was shown to exhibit gene-environment interaction and this was used in the case to reduce the conviction from murder to manslaughter.
The jurors in the case had a genetic essentialist view i.e. our phenotype is set by our genotype, and it and its effects are out of our control.
One issue: people with this background did have increased rates of violence but most of them were not guilty of committing violent crimes (and the gene might be pertinent to impulsive violence and to premeditated violence).
We have to make a distinction between an individual responsibility for developing a disease vis-a-vis his or her responsibility after that.
Phenylketonuria is an example – they are not responsible for developing the disease but they can do something about managing the disease.
There is a campaign to de-stigmatize mental illnesses/health but saying it is not the responsibility of the individual as it is genetics. Track record is mixed:
Positive: reduce blame somewhat;
Negative: viewed as more dangerous.
Do people who know their genetic information e.g. risk of disease, behave differently? Track record is mixed too, usually:
Less confidence in psychosocial interventions (therapy);
More confidence in biological interventions (e.g. medicines).
Considered one of the leading pioneers/scientists in the field – did the first study on behavior genetics studies of personality.
Discussion of history, major discoveries, promising technologies, epigenomics etc.
Talked about endophenotypes i.e. intermediate phenotypes not visible to the naked eye, needs some processing e.g. brain imaging, functional MRIs.
Genetics might be like where statistics was i.e. it will come incorporated into a lot of fields as it is a very powerful toolset for understanding behavior, and it can be coupled with imaging and neuroscience.
There is still a debate on how informative one’s genetic sequence is, partly because the difference between the biometric heritability (through twin studies) and the % variance accounted for by GWAS is still huge (i.e. missing heritability):
BMI: 50-80% of the former vis-a-vis 1.5% of the latter;
Height: 16% vs 70-85%;
Serum urate: 6.8% vs 40-70%;
Total cholesterol: 12.4% vs 50-75%;
C-Reactive protein: 5.0% vs 40-60%.
Mental health disorders all below 20% for the latter.
Epidemologists use a statistic called area under the curve to predict disease outcomes, between
0.5 = 50% i.e. chance like flipping a coin; and
1.0 = 100% i.e. perfectly accurate;
0.75 = 75% makes it useful for screening populations to make decisions/diagnosis.
Some might want it to be 99% but that might be very high
Most disorders, except for macular degeneration are below 75%.
Reasons to be optimistic despite the above:
Rare variants mean for some, the genetic info is already predictive e.g.
30% of individuals with a particular gene will develop schizophrenia;
Almost all with another particular gene will usually have autism;
Rare mutations in 3 genes lead to Alzheimer Disease;
1 out of 25 couples share a rare recessive variant that raises the odds of a child with severe congenital disorder
More variance are found with each GWAS study (as GWAS samples become larger) and it is predicted that eventually we can account for all missing heritability (it might take 50-60 years).
This will always be limit because phenotypes are not perfectly heritable;
There will also always be some statistical inaccuracy.
The above complicates the decision making based on genetic prediction.
We can already identify subgroups with very high or low disease risk.
e.g. high number of risk alleles and odds of schizophrenia.