BSCI222 – Lecture 15 (10/24/13)- Chapter 24: Quantitative Genetics- Mendellian genetics has discontinuous characteristics; here we have continuous characteristics, where the population has a bell curve for the trait, with a wide range (continuous distribution of phenotypes).o Why would you have this? Might have: multiple genes affecting the trait, or multiple alleles at a locus, or interactions among genes (epistasis), or environmental effects on the phenotype. o If you plant many cuttings from the identical genotype, they will not all be the same in phenotype (environmental variation, such as water availability, etc.)o One locus with 2 alleles (in equal frequency) -> get 3 genotypes (AA, Aa, aa) -> same bell curve as genetically identical, but more spread out (wider). o Many more alleles: much broader distribution of genotypes, and have a continuous trait curve when including the environment.- Partition the variants (partitioning the phenotypes):o V_G (variation due to genetics) and V_E (variation due to environment). Heritability (h2, in a broad sense (h2B, is equal to V_G/V_P (what proportion of the variance is due to genes and what proportion is due to the environment). - Example: 2 alleles, 2 loci, additive effects on the phenotype:o 2 pathways to producing purple pigment in a wheat plant. A wild type or B wild type adds pigmentation units (2 different genes, with additive effects). F1 hybrid -> interbreed -> F2 (TEST: know how to do this table with wheat colors) Purple is 1/16 in F2. More loci -> more columns, bell curve starts appearing. - Even with just 2 alleles, with a lot of environmental interference it becomes impossible toknow the genotype from their phenotype. (Example: height in people, tall dwarf or short normal?)- Analysis of variance: sum of squared deviations from the mean. Greater variance means abroader curve. - Components of Genetic Variance:o V_G = V_A (additive components) + V_D (dominant components)o Perfectly additive: adding one allele gives you an exact phenotype (didn’t hear)? bb = -a, bB = 0, and BB = +a Usually don’t have the case of it being perfectly additive. bB’s distance from 0 is called d. If the trait were perfectly additive, then the heterozygotes should fall right on the line between the two homozygotes. If the heterozygote is above the line, this indicates some degree of dominance (amount that it is above the line is d).  Why is there a distinction between additive and dominant genetic variance?- Maybe heterozygotes grow faster; a farmer/breeder will select for it. Keep selecting heterozygotes -> will never get any improvement, always 1:2:1. Cannot breed for dominance. Can, however, select for the additive components, which will not leave you stuck. Actually selecting for BB will give you a higher mean for the population overall. Narrow sense heritability is what breeders care about: h2N = V_A/V_P- How do you know your breeding will give you genetic gains? Must somehow estimate what improvements you will get.o First, must get estimates of heritability: will the trait respond to selection?o Also, need estimate of how hard you must select in order to get that gain. If you know what kind of response you expect, can calculate the potential economic benefit.o Heritability is typically estimated from offspring of parent regressions: make a bunch of crosses among the adults in your population, plot the average of the 2 parent phenotypes (called midparent value) (i.e. one big one small, = ½. Two juicy parents = 1) against the phenotype of the resulting offspring. Crazy points everywhere: low heritability (or none), no relationship. If there is a positive correlation: must have additive genetic variance, for which you can select. The heritability in the narrow sense is equal to the slope of the line (steep slope means strong heritability). If heritability for one trait is higher than the heritability for another, then you can change the trait with higher heritability faster (can get bigger cows faster than cows with higher milk yields). Traits linked to reproduction do not have much additive variation left (or any genetic variation), because evolution has selected for it for so long.- Heritability estimates are not absolute.o Depend on the particular population and environment of study, and sensitive to exact methods used to measure the trait (i.e., how long are you measuring growth?To 5 weeks? To 10 weeks? Did you use a ruler correctly?).o Different populations have different allele frequencies to start with, thus having different heritabilities. Extreme case, if population is fixed for a locus, will have zero heritability (no room for improvement).o Genotype that works well in a wet environment will not work as well in a dry environment.o So why do it? Indication of response to selection we can expect (funding). Typically good investment, because it is a one-time cost to change the population permanently; gain will be in the population forever, giving the benefits for a long time. Like improving the software. Margins are of small percentages; an organismgrowing 2-3% more efficiently will take over the market. - Another measurement, is to take a group of parents that have a slightly higher mean phenotype, from the population. Offspring then have a slightly higher mean phenotype than the average population.o Called the response to selection, is another measure of heritability.o R = h2 * S (the selection applied), or h2 = R/S (how far does the population move after applying some amount of selection). - Limits to selection?o Continuously select -> get best locus across the genome? Run out of genetic variation? Is Clifford possible?o Might be possible, but not in every case. Appears that there is a selection limit, past a certain point. Thoroughbred horses have hit a limit for increased speed. Experiment on corn oil content, began in 1896, still continuous improvement (the low-content has stopped because can’t have less than 0% oil, called a phenotypic limit) (high-content has so far seen a difference of 32 SD’s). Limits in number of bristles on drosophila. Why the difference? The more genes you have, the bigger target for mutations to happen and selecting on new variations that weren’t present before. o Starting with a smaller population makes it more likely to hit a selection limit (starting with less variability) (exhaustion of genetic variability).o Essentially, selection limits are not a problem for