BSCI222 – Lecture 16 (10/29/13)- Quantitative Genetics, part 2: Your Genetic Identity- The biggest challenge for Biology in the next century is determining how to take the information of genotype that we obtain, and then determining phenotype from it. - Personal genomics – what can we learn, when is it useful, when is it a waste of money or dangerous.o TED video: AVPR1a (cheating locus RS3 repeats, allele 334, has strong correlation to partner bonding scale, men only) And DRD4 in dopamine. - Prenatal Testing: (homework paper) before this paper, had to do invasive sampling. Now just filter mother’s plasma, and pick up the DNA being sloughed off by the embryo and going into the bloodstream. Won’t be long before this technique can be used to sequence the fetus’s genome before birth. o CEO of one of these companies refuses to sequence children who aren’t sick; saysit’s not just for one and doesn’t believe parents have a legal or ethical right to this information.o CEO of 23andMe got both her kids immediately tested when she could, mostly for Parkinson’s. Does she have a right to that information about risk?o American Academy of Pediatrics has come out against genetic testing of children for adult-onset diseases. But parents want it. Should they be allowed? That’s an ethical decision.- So what do we know about genotype -> phenotype?o Relatives look alike because of shared environments (pre and post natal) and shared genetics (additive components and dominance components). o Twin studies (can do twin 1 v. twin 2 comparison graph, just like aforementioned offspring to parent comparison) Even identical twins have a lot of variation in environment, depending on when the separation occurs: before implantation (separate amnions single chorion and placenta) or after (single amnion, chorion, and placenta OR separate amnions, chorions, and placentas). Dizygotic twins share a placenta and might share a chorion (amniotic sac). Difference between twins is estimate of additive and dominance components. Dominance component is actually pretty small here. Heritability estimates from twin studies: finger ridge counts are 98%, height is 66%, and IQ is 34%. High heritability, so should be able to find the underlying genes, right? Not necessarily; high heritability can result from segregation of hundreds of loci, making it very difficult. o Genome-wide association studies: gather many people that vary in the trait, and look for associations between genotype and phenotype. Human height; look for mean height of people at each genotype (if it’s statistically significantly different, found an association). Genotyped 500,000 snips from 13, 665 individuals; found 20 snips that were associated with height, explaining only 3% of the phenotypic variance. How could such a big study have found so little? Another paper, used 2 million snips from 183,727 individuals. Found at least 180 loci influencing height, explaining about 10% of the variance in height. These are the best we have so far, and Dr. Kocher sees them as complete failures; not explaining much of the variation. Different study: used 294,831 snips, made statistical model that explains about 45% of height variance. Turns out, height is highly heritable but also highly polygenic. Not by a small amount of genes with a huge effect, but LOTS of low-frequency alleles, each with a very small effect. The strongest alleles have an effect of about 0.25 inches.o The genetic population today means that most of the diseases we care about are being controlled by rare deleterious mutations that have occurred relatively recently.- Is intelligence genetic?o IQ tests Many kinds of intelligence, but scores for different kinds of intelligence are highly correlated. IQ scores are subject to a variety of cultural testing biases/errors. Average IQ scores have been rising about 3 points per decade, but population can’t be changing its genetics that fast, must be some other effect. IQ is not a perfect measure of success. Social effects: - Birth order (coming later means lower IQ, but if the older siblings die your IQ jumps up to their level) (thought to be because of first-born getting more attention, and if the sibling dies you get a richer environment to develop in). - Maternal Environment: major portion of brain growth occurs in utero. IQ is correlated with brain weight. Twins are lighter at birth and typically have 4-7 points lower IQ. Dietary supplements, smoking/alcohol, lead. The IQ scores of mono-chorionic twins are 12% more similar than di-chorionic twins; the one extra membrane during development increases the environmental variance. IQ is highly heritable: broad sense about 48% (VG/VP), narrow sense about 34% (VA/VP) Genetic, so can we find the genes? Another huge wide association study; after many phases, and correcting for multiple tests, only found 1 significant association, 6 snips, showing a correlation with IQ of 0.11. Again, highly heritable trait but can’t really find the genes responsible for it.- So what will sequencing your genome tell you? Probably not much about height, or IQ, or diseases, or most traits. Most traits are quantitative and controlled by a huge amount ofgenes and not really tractable for predicting this relationship between the genotype and phenotype. The exceptions are Mendellian-type diseases, which were found pretty quickly after the human genome project.- Race and Intelligence (James Watson’s faux
View Full Document
Unlocking...