Lec 15

• Breeding Value (BV)


  Genetic merit of an individual as a parent
• Estimating Breeding Value (BV)

  1. Sum of additive genetic values of all its alleles

  2. P = G + E

  3. P = A + D + I + E

  4. A (additive genetic effect) is the transmittable part
• Phenotype value (P)


  Own performance, determined by Genotype value (G) and Environmental effect (E)
• Genotype value (G)


  Joint effect of all genes in all relevant loci
• Breeding Value (BV)


  Sire's BV + Dam's BV = Progeny's BV, but not all genes are transmitted due to Mendelian sampling
• Estimated Breeding Value (EBV)


  Actual BV is unknown, estimated from own performance and performance of relatives
• Estimating EBV


  1. EBV = ½ BV sire + ½ BV dam

  2. EBV = ½ EBV sire + 0 (EBV dam = 0 for random mating)
• Value of information of relatives

  • The closer the relative, the more valuable the information in estimating BV
  • More relatives and closer relatives increase confidence in EBV estimate
• Selection decision


  Select for additive genetic effects (A = BV), as dominance (D) and interaction (I) effects are not transmittable
• Average effect of alleles (a)


  Mean deviation from population mean of individuals receiving that allele
• Dominance deviation


  Difference between genetic and breeding values, due to dominance effects
• Breeding Value (BV)


  Sum of average effects of alleles on all relevant traits of an individual as a parent
• Genotypic value of offspring


  Determined by average effects of parents' genes, which cannot be directly measured
• Breeding Value (BV)


  Parental genetic value transmitted to offspring, can be estimated from ancestry, individual, and progeny records
• Relative importance of ancestry, individual and progeny records in determining an animal's breeding values for milk production traits
• Estimated Breeding Value (EBV)

  Calculated as deviation from population/herd mean, using individual record and records of relatives
• Accuracy of EBV estimates


  Measure of confidence that EBV reflects true genetic merit, improved by more information from relatives
• Estimated Progeny Difference (EPD) for body weight of two beef cattle bulls


  • Bull A: Birth wt +4, Weaning wt +20, Yearling wt +30

  Bull B: Birth wt -2, Weaning wt +5, Yearling wt +20
• Differences in phenotype among offspring of the same family are due to segregation and independent assortment of genes/alleles, and environmental effects
• EPD


  Estimated Phenotypic Difference
• Traits for beef cattle bulls

  • Birth weight
  • Weaning weight
  • Yearling weight
• Bull A


  • Birth weight: +4
  • Weaning weight: +20
  • Yearling weight: +30
• Bull B


  • Birth weight: -2
  • Weaning weight: +5
  • Yearling weight: +20
• • Difference in birth weight: 6
  • Difference in weaning weight: 15
  • Difference in yearling weight: 10
• Estimating BV Using Own Record

  1. EBV = h^2 (P - P_)

  2. Accuracy of estimate = √(bg) = √(h^2 (1)) = h

  3. b = h^2

  4. g for individual = 1 (for relationship with itself)
• Accuracy of the BV can be improved if we measure the performance of its progeny and sibs
• Predicting EBVs


  1. Measure the phenotype >>> Phenotypic value of the individual

  2. Example: Bull A with a yearling weight of 250 kg, what is its BV?
• Estimating EBV Using Own Individual Record

  1. EBV = h^2 (P - P_)

  2. Accuracy EBV = √(h^2 (1)) = h
• Within family variation


  Differences in phenotype among offspring of the same family, due to segregation and independent assortment of genes/alleles in gamete formation and environmental effect
• Using EBVs


  1. If the male parent is mated to a large number of breeding females (average BVs = 0), then:

  2. EBV of male parent = h^2 x phenotypic value

  3. EBV of female parents = 0

  4. Expected genetic value of offspring = (EBV male parent)/2
• We do not expect all offspring of the bull (EBV bull = +7 kg) to be +3.5 kg above average because the genetic value expected for the offspring is an average value. Some offspring may be > +3.5 kg and some < +3.5 kg.
• Repeatability (r)


  Telling future performance from 1st or 2nd record
• Estimating EBV Using Repeated Records/Measurements

  1. EBV = b (P - P_)

  2. b = nh^2 / (1 + (n-1) r)

  3. n = no. of repeated records

  4. r = repeatability

  5. Accuracy of EBV = √(bg)
• Additional records of relatives and repeated measurement are most valuable for traits with low h^2 and low repeatability. Each additional record will improve accuracy of estimate, but each additional record contributes less to improvement of estimate and accuracy compared to previous record. Using many records of relatives may also lengthen generation interval.
• Full-sibs


  Individuals having both parents in common
• Half-sibs


  Individuals having a common sire
• Full-sibs are expected to have 1/2 of their genes in common, and half-sibs are expected to have 1/4 of their genes in common.
• Estimating BV Using Sib Information

  1. Intra-class correlation (t) = correlation among sibs

  2. t_FS = 1/2 h^2 + c^2_FS

  3. t_HS = 1/4 h^2 + c^2_HS

  4. c^2 = environmental correlation among sibs (often ignored)

  5. b = (gh^2 n) / (1 + (n-1) t)

  6. g = 1/2 for full-sibs and 1/4 for half-sibs
• Accuracy always increases as n or number of sibs increases though the increase is less with each additional sib record: each additional sib contributes less than the previous. Accuracy is higher for traits with higher heritability. Sib information is useful for traits with low heritability. With low heritability and environmental correlation zero, full sib records contribute as much accuracy as individual record.
• Accuracy of EBV using sib information will increase as n increases though the increase is less with each additional sib information. Accuracy is also higher for traits with higher heritability. Sib information is useful for traits with low heritability. With traits of low h^2, full-sibs information contributes as much accuracy as individual record. Use of sib information does not alter generation interval. Useful in evaluating carcass traits. Never result in high accuracy (limited to 0.5 or less and full-sibs are hampered by common maternal environment). Use to supplement on individual own information.