Quantitative Genetics Flashcards

Question

Findiong Vp

Answer 1

Gene and envirnemnt affect Vp in adidtive way --> Additive relationship between genes and envirnment (easy to comvine genetic and envirnmental varaince) Vp = Vg + VE - Vg = variance effects due to genes - Ve --> Varaince due to envinment VGXE = can be determined too BUT that seperate source of variance also forms an additive component of Vp and is often lumped in with Ve - VGXE = another aditive effect --> also adds in additive way to total phenotypic variance Vp = Vg + Ve = VGXE

Answer 2

NS can only act on heretible varaition to cause chnage in one generation to the next - Whether something is heritable or not is not just yes or no --> genes are heretible BUT envirnment is not heretible --> NEED to quantofy partial heretible effects to know how phenotype can change due to NS - The combination of Bg and Ve results in partial heredity Heretible varaition = part of Vp we are concerned with

Answer 3

0 -1 0 -- there is no relationship between parent and offsrping phenotype - Offspring are not correlated to parents - All envirnment -- no varaince due to genetics 1 -- Offspring phenotypes are expactlet like parents - Phenotype is expactley like parents -- if know parent know offspring phenotype

Answer 4

Broad sense heretibility H^2 = Vg/Vp Proportion of total phenotypuc variance (Vp) made up of genetic varaince - Proportion of total varaince that is ascribed to genetic varaince Narrow sense -- proportion of phenotypic varaition that is passed between parents and offspring in a straight foward way (easily predictable way) - due to additive affects - Proportion of all genetic varaition due to additive affects h^2 = VA/Vp

Answer 5

Broad = encompasses all of the gentic input to the trait - Ecompasses all potential genetic affects acting on phenotype - All varaition due to any type of genetic affect - Includes genetic affects that it is hard for NS to act on - Braod = not measure of heretibiulity critical to making predictions H^2 = Vg/Vp Narrow sense -- proportion of phenotypic varaition that is passed between parents and offspring in a straight foward way (easily predictable way) - due to additive affects - Proportion of all genetic varaition due to additive affects - Part that matters --> component of Vp that is due additive h^2 = VA/Vp

Answer 6

Vg = can be broken down into different components - Can break down genetic variance Vg = Va + Vd + Vi Va = additive genetic varaince Vd = dominance genetic varaince Vi = epistatic interaction genetic varaince ***Not hard to measure them independently - Hard to get in nature

Answer 7

Va = drives narrow sense heretibility - Affects parent offsrping relationshio Vd and Vi = interfere with direct parent phenotype to offspring phenotype relationshio - Interfere with some of parental contirbution to your phenotype = interfere with direct realtionship - they’re difficult to measure without knowing the genotypes. - More importantly – they interfere with the straightforward prediction of the effects of natural selection

Answer 8

Hard to do directlet --> hard to fo without finding h^2 Don't know Va without knowing h = we measure h^2 directley and then get Va

Answer 9

Heretibility involoved the correlation between parent and offsrping phenotypes --> most straightfoward way to get h^2 = through midparent regressions - easy way to get h^2 = to compare parent to offspring phenotype using mid point regression

Answer 10

Take parent average phenotype - Expect the offsrping the match the average of the parents --> use expectation to see gow well data matches - Estimates h^2 h^2 = equal to the slope of a linear regression bteween the midpoint phenotype (average of mother and father) and the offsrping phenotype (or mean offsrping phenotype for multiple sublings) Overall in regression = looking at average parent vs. Avergae offspring

Answer 11

Seeing how Y=mx + b fits varaibles --> finding linear relationship between varaibles NAME = has to do with motivation to do this -- has to do with heretibility - On avergae children = look more like popultion mean not like parents = regression to popultion mean --> regression away from parental values and towards popultion - Complete regression = no heretibility - No regression = parent phenotypes are like offspring phenotypes

Answer 12

Overall in regression = looking at average parent vs. Avergae offspring Left = NO statistical coreelation --> parent phenotypoe dies not predict offspring phenotype - h = 0 Middle = Have trivial amount of heretibility - Parent phenotype predicts something but not completley - h = 0.5 Right -- h = 1 Value of heretibility = value of regression line - Slope = 0 --> like popultion mean not parent --> h = 0 - Slope = 1 --> h = 1 --> all varaince driven by additive genetics - h = 0.5 --> 1/2 varaince due to parent genotypes; 1/2 variance due to envirnmnetal non-additive components

Answer 13

1. Height -- found h^2 = 0.84 MEANS 84% of varaince in height = due to genes inherited by parents 16% = due to others SHOWS many genes involoved + strong phenotypic affects of strong straight foward parent vs. offspring 2. Looking at behavior in spiders -- looking at how social they are - Looked at how the spiders were together Results: Y = 0.66X + 4.47 - h^2 = 0.66 -- 66% of phenotypic variance due to additive genetic varaition ***Looking at parent traits vs. offsrping traits --> plot --> find h^2

Answer 14

In the real world we might not have perfect data for mid-parent regressions --> it is often hard to know who the father is -Sometimes hard to know parent sitution (might not know who the father is) INSTEAD -- we can measure h^2 using a single parent/offsrping regressions (can do it based on only mom and offspring = single parent regression) - Single parent/offsrping regression = under estimate heretidity by exactleyt half --> THEN h^2 is equal to 2X the slope in single parent regression - Only accounts for 1/2 genetic varaition = underestimate h by extacley half --> h^2 will be the slope X2 (because know underestimate by exactly 1/2)

Answer 15

1. Might be hard to get data for mid-parent regression 2. Parent and offsrping are likley to share envirnmental varaition - parents and offsrping share envirnment that other members don't share not only genes - Shared components of envirnments vraiation that conturbutes to parent vs. offsrping relationship pairs = complicates h^2 - Can be envirnment rather than genes Example -- shared envirnment -- might share where they live --> like if they gave a lot of food in area -- parents get the food + kids get the food - Experiment -- forced bird pairs to raise foster bids -- can takie into account envirnmental varaition to rule out envirnmental contribution of varaibility -- canges slope of regression h^2 = the upper limit of hereteibility in popultions --> because can be due to envirnment rather than genes = h^2 measured in mid-point regression = shows the upper limit of varaibility because some amount of slope is because of shared envirnment

Answer 16

1. Sibilings vs. half siblings --> Sibling studies predicted on known coefcient of relatness (r) 2. Twin studies -- humans

Answer 17

coefeciant of relatedness = r Traits = nested 1/2 sibling design - Have males (Sires) and females (Dams) Have males = multiple females = get different sets of offsrping -- get full sibling and half siblings - Differences in degree of genetic relatedness --> Gives null (compare to data to see Va) Full sinlings = X2 similar vs 1/2 siblings --> Using ANOVA --> ANY PHENOTYPIC VARAINCE IN HERETIBILITY --> can produce affect Va from other components of varaince - Uses a nested ANOVA framework to parse apart components of variance. VA equals four times the phenotypic variance among half sibling families Overall: gives us ways to break Va down Issues: 1. need organisms you can get into design 2. need 30 males --> each mate with 3 females --> 270 offsrping --> very statistical = need big sample size - works well with small organisms (can many in lab) 3. Hard to do with many wild animals

Answer 18

Differences in degree of genetic relatedness --> Gives null (compare to data to see Va) Full siblings = share 50% of genes -- 50% chance siblings has alleles (50% allele is shared) - r = 0.5 1/2 siblings = share 25% of genes --> can share more genes from mom or not more - r = 0.25

Answer 19

Done in humans --> compare identical (monozygotic) to fraternal (dizygotic) twins - can do mid parent regression but not other routes - Take advantage of known relationship in coreficants of relationships in groups of indviuals - Compare mono vs. dizygotic twins --> how much phenotypic varaition between mono and how much phenotypic varaition between dizygotic Mono = share all genes --> Share same genes + Samne envirnment - r = 1.0 Dizygotic = share 1/2 of their genes --> Same relatness as other siblings BUT experince the same envirnmental varaintion (similar to how mono share envirnment) - r - 0.5

Answer 20

Prameters = refer only to the specific popultion (genetic makeup) in a specific envirnment - Specific to that envinment --> if put popultion in new envirnment = change Ve = chnage h^2 - Specific to popultion AND specific to popultion in that envirnment that it is in -- only for popultion in that envirnment When we measure heretibility it is only value within that popultion -- it is useless for making comparisons between popultions

Answer 21

Interferes with ability of offspring/parent phenotype

Answer 22

if we know heritability of trait = we can predict how NS will affect it - Need to know h^2 to know what will happen to trait in next generation In order to know how something will change over time --> need to know h^2 -

Answer 23

We can predict a particular strength of selection using Breeder's equation

Answer 24

R = h^2s R = response to selection in the offspring generation - Phenotypic response --> change in mean trait value from one generation to next due to h^2 S = selection differential in the parental generation - Select subset from parental generation to S/R -- get difference between overall mean and selected mean = get S Response to selection = due to strength of selection + heritability of trait h^2 -- not actually squaring something -- just put squared to differentate from other times h value is used (Called breeder's equation because used in breeding in plants)

Answer 25

restriction on selection -- limited becaused h^2 is not perfect -- not 1 Response to selection = due to strength of selection + heritability of trait

Answer 26

Selection differential - Difference in traits mean in parents vs. the trait mean in the selected individuals ***S = within overall population mean In situations where selection is based on survival (or choice for reproduction in artificial selection program) -- S equals the difference between mean trait value of survivors (or individuals selected to breed) and the overall mean of the starting population Easy to see in breeding programs -- choose who survives/reproduces

Answer 27

Experiment Blue + Red = overall population -- mean trait value Blue = Red = ines that reproduce to impose selection difference between blue mean and red mean Artifical selection in muce tail length t/ = mean tail length of the population of mice t* = mean tail length of the mice chosen to reproduce Difference between t/ and t* = S

Answer 28

Gall example White arrow = mean gall diameter of the population - White = mean all populations Black arrow = the mean of the survivors - Black = mean popultion of gall that survive = only ones that reproduce ***Have a binary on if you survive or not Distance between the two = S

Answer 29

The relationship between the trait value and the relative fitness - Scalable into complex situations More flexible way to think about strength on phenotype Can account for continous fitness varaition + varying fitness varying as a function of the trait - Look at fecindity -- relationship between varaibles = selection gradient ***Acts in the breeder's equation in the same way

Answer 30

Can use selection gradient to apply our quantitative genetics tools to more realistic complex fitness situation Can use technique in more complex settings

Answer 31

Differential = trait as binary survive or not survive - Just have number of individuals Selection gradient = Continuous relative fitness - Slope = selection gradient Can convert between the two

Answer 32

Flowing time Relative fitness = number of seeds Earlier flowing time = increase # of seeds = increase fitness - Mathematical relationship = selection gradient

Answer 33

Selection gradient = do not need to be linear

Answer 34

Response to selection -- change in the population mean from one generation to the next

Answer 35

R < S because h^2 is less than 1 - The response in selection R is smaller than S because h^2 is less than 1 Response on selection = sub standard based on h^2

Answer 36

Look at Mid parent regression -- parent phenotypes + offspring phenotypes -- same plot used to get h^2 - Look at distance of parental phenotype and distance of offspring phenotype X Axis -- push trait (pushing parent trait) by X much and see chnage in offspring (See how much Y chnages) Graph -- pushing P --> P* and seeing chnage in Y value (Change in offsrping trait value) - When look at slope -- look at how likley offspring match parent (slope is h^2) - Look to see for X push on X = get Y (get that much increase) in offspring - Run across mid parent regression that rise along using h^2 - See what happens if increase parent trait by X much -- see chnage in Y value -- change in offspring h^2 = slope

Answer 37

Galls -- we are solving for R -- how R relates to the parental population Mean gall diametere in the popultion is 25 mm (overall parental pop. mean) h^2 = 0.18 --> 18% pf selection is due to genetics Survivors = 21 mm R = h^2s S = 21 - 25 = -4 (moving down) R = 0.18 X (-4) = -0.72 New mean of gall size = Starting + R = 25 + (-0.72) = 24.28 - Comparison if to the parental mean (25 NOT 21) --> impose R on the overall population mean not the selection mean Overall: based on h^2 not much shift --> have some shift but not so much considering selection - Don't have huge shift because a lot that links parent and offsrping is not genetic ***Reall

Answer 38

Really just looking at rise over run for change (-4) = run -- look at how much rise there is - S = run - R = Rise 18 = 25h^2 h^2 = 25/18 --> h^2 is slope = rise/run

Answer 39

Here = Positive R = adding to parental value

Answer 40

Selection for larger trait value = Positive S Selection for smaller trait value = Negitive S S = gives directionality

Answer 41

We can also use equation to estimate the heretibility from artifical selection experiments - Can run selection experiment and use Breeders to calculate h^2 --> Might want to know if there is Va to select in the future Example -- If running an agricultural breeding program to imrpove corn - Mean percent popped = 60% - You think there is a lot if varaince in phenotype and you want to know how much is genetic -- want to know how much you can change trait - Breed -- take from popultion with 85% popping GRAPH -- parental vs. Offspring - Offspring increase popping rate (60% --> 78%) - Slope = h^2 -- get slope with rise/run - Positive slope because increase trait value - S = run - R = Rise For 0.6 --> 0.85 increase in X (increase in S) = get 60 --> 78 increase in Y (increase in R) 18 = 25h^2 h^2 = 25/18 --> h^2 is slope = rise/run

Answer 42

18 = 25h^2 h^2 = 25/18 --> h^2 is slope = rise/run h^2 = 0.72 25 -- 85 - 65 18 -- 78 - 65

Answer 43

S = selected mean - starting mean - Gives directionality R = offsrping mean = starting mean - NOT using selected mean S = gives directionality R = gives response

Answer 44

Know heretibility without knowing underlying genotypes

Answer 45

NEED TO BE CAREFUL -- h^2 represtns the upper limit of heretuibility + other varaibles that are hard to break down - Over estimate h^2 because of shared envirnment (looks like heretibility but heretibility but really shared parent offsrping envrirnment not shared genes) Example -- Birds that are in area (area might be warmer) = get more sunlight = birds in area are faster and bigger BUT they share the trait because of shared envirnmnet - Outcome is due to envirnmental similarity - Also applyes if have same diet in parent and offspring

Answer 46

1. Maternal Effects 2. Epigenetics

Answer 47

Envirnment of mom affects not gene - Resemble mom but resemblance is due to envirnmnet that shapes aspect of development Example -- Egg size is based on food mom had not moms genes -- due to moms envirnment not genes BUT different moms have different envirnments thats leads to difefrences that look genetic

Answer 48

Trangenerations epigenetics -- Trangenercation effects - opperates in the lab --> studies on lab organisms now studied in humans Affects that we are learning more about OVERALL -- ssRNA for a gene regulates or RNAi regulates -- the effect of RNAi is seen after generation after intial generation - Aleter gene expression that gets passed down to offspring not just in the 1st generation - Get connection between parent vs. offsrping dur to epigenetics not heretible changes

Answer 49

NOT trivial KNOW there are some issues with measuring additive genetics varaiation in human traits

Answer 50

Nutritien and transgeneration effects Experiment -- knew pedigrees + relationships + Knew there had been major famines Epidimeologists looked at the health outcomes in grandchildren of people who went through the famine Results: Strong effects + complex - Found difference between men/women -- difference on if a man/female went through famine = difference in sex of grandchildren IF a Males = less usbstantial effect if grandfather went through famine If Females -- more substantial effect if grandmother went through famine SHOWS envirnmental is complex -- heretibility is very complicated --> effects exist in huamns but very complex + hard to predict - Very complicated -- difefrent sexes are very different

Answer 51

Envirnmental -- NOT always straight fowards envinmental effects Example -- Maternal effects and transgenerational epigenetics effects on human phenotypes may be more common that previously thought

Answer 52

ALL complicated our ability to quantify genetic components of traits within human popultopms - Complicates ability to make inferences within popultions we measure -- we often overestimate heretibility Measuring narrow sense heribility = far from trivial --> espcially complex traits in humans because can't do the same experiments in humans Even when we feel we can draw string conlsuioons about heretibility we are still limited to talking about the role of genetics in THAT popultion - Strong heretibilility parameters = tell us about proportion of Va in popultion in THAT envirnment --> only have how much genetics drive phenotype in that envirnment - Doesn't say anything about between popultions --> hard to look at heretibility across different asopects of human society

Answer 53

Even when we feel we can draw string conlsuioons about heretibility we are still limited to talking about the role of genetics in THAT popultion - Strong heretibilility parameters = tell us about proportion of Va in popultion in THAT envirnment --> only have how much genetics drive phenotype in that envirnment - Doesn't say anything about between popultions --> hard to look at heretibility across different asopects of human society Example -- You are intersted in imporving beed yeilds through selective breeding Start = go to different cattle in popultion and look at phenotypic varaition in beef yeild in cattle to get an idea of Va - Get idea of genetic differences in popultion to leverage in experiment THEN -- you gi to a couple of large farming operations to measure additive genetic varaince IF in popultion 1 -- mean mass = 450; Vp = 10,000 IF popultion 2 -- Means mass = 850; Vp = 10,000 Difference in mean BUT similar amounts of phenotypic varaince GRAPH (off vs. Parent) -- same slope but one mean is 450 (lowser) and other mean is 850 (higher) - Results --> heretible trait -- strong relationship - h^2 = 1 in both popultions = within popultion all of the varaition is driven by Va (high heretibility but big difference in trait values) - Both slopes = 1 = h^2 = 1--> have 100% Vp explained by Va - If select to imporve beef yeild --> is one popultion genetically bettwe fir beef yeild (if H^2 is 100% in both) -- is there genetic difefrence between popultions On the surface = seems to be geentic difefrence but the data does NOT suggest that Have we found genetic difefrence? h^2 us high = explained by Va --> tells us little about envirnmental varaince within popultion --> genetics could be the same but both popultions are in difefrent --> if foo is good in one and bad in other --> the differebnce is due to envirnment We don't know if Va is what is difefrent between popultions

Answer 54

Compare humans popultions --> diferent traits value parameters don't tell you difference in genes We don't know if Va is what is difefrent between popultions Human popultions = not in homeostasis --> have lots of envinmental variation - Envirnmental varaition can play out in one generation or across generations Envirnmental varaition makes saying things about genetic differnces meaningless If people say heretible and differnces is genetics ---> They are using a misapplication of biology Each comparison only works for varaition in popultion you are measuring in that time

Answer 55

Murray -- pepetuating ideas of human groups based on heretibility Wrote about bell cirbe about intelligneve as heretible trait + distrubution in society leavergaed things about heretibility of traits and implying ideas are determinaistic --> idea that certain outcomes if IQ is not envirnmmetal driving then it must be due to genetics = not our probelm to fix Lots of varaition varied with envirnment that within and btewen popultions --> just because intelignece can have heretibility doesn't mean they are hardwired into genes It is easier to change envirnment --> faster than giving up Realization of QG --> closed idea of taking Murray seriously - He would reducve the amount of varaition (if only genetics) = bad + have huge envirnmental components + affects societal outcomes without racist outcomes

Quantitative Genetics Flashcards

(80 cards)