Predicting Genetic Gain

1
Predicting Genetic Gain
2
Predicting Genetic Gain

• Selection Theory
• Why does selection work?
• Taking the best phenotypes improves the
  population genetically
• Changes allele frequencies
• Prediction of genetic gain.
• Enhancing genetic improvement!!!

3
Predicting Genetic Gain

• Breeding Value (BV) The value of an animal as a
  (genetic) parent.
• Independent Gene Effect The effect of an allele
  is independent of the effect of the other allele
  at the same locus (dominance) and the effects of
  alleles at other loci (epistasis). Additive
  effect.
• Breeding Value The part of an individual
  genotypic value that is due to additive effect
  and therefore transmittable.
• Estimated Breeding Value (EBV) A prediction of a
  breeding Value.

4
Progeny Differences

• Progeny Difference (PD) or Transmitting Ability
  (TA) Half of an individuals breeding value. The
  expected difference of the individuals progeny
  and the mean performance of all progenies.
• Expected Progeny Difference (EPD) or Estimated
  Transmitting Ability (ETA) A prediction of a
  progeny difference.
• Additive Gene Effect Independent gene effect.
• Additive Genetic Value Breeding Value.

5
How to Maximize the Rate of Genetic Gain

• Which Traits to select? (Primary / Major
  question)
• Should I select based on correlated traits?
• Should I selected many female replacements or
  just a few?
• Should I use many males or just the very best
  ones?
• Should I use well-proven, older males or
  promising young ones?
• Should I base selection on individual performance
  or should I consider information on relatives?
• Should I select strictly within my own herd or
  flock or should I look to other populations for
  replacement?

• These questions arent easy to answer!!! Understa
  nding the factors that affect the rate of genetic
  gain can help answering these questions! It can
  help us to develop selection strategies and
  design a breeding program!!

6
Predicting Genetic Gain
7
Predicting Genetic Gain Key Equation
8
Predicting Genetic Gain
The rate of change on a population under
selection. Difference between parental generation
and offspring genetic values
Related to the proportion of the selected animals
to became parents.
Measure of the phenotypic variation on the
population.
Proportion of the phenotypic variation that is
inheritable.
The amount of time required to replace one
generation with the next. Average age of sires
and dams.
9
Predicting Genetic Gain
10
Predicting Genetic Gain
Proportion Selected
Selection Intensity.i
Selection Differential (in Parents).S
Variation.sp
Response per Generation?G
Heritability.h2
Generation Interval ..L
Response per Year . ?G/year
11
Predicting Genetic Gain
?G change in population mean per generation due
to selection ?G ?gen1 - ?gen0
12
Selection Differential (S)
Selection Differential is the superiority of the
selected animals in relation to the population.
Is the difference between the average of sires
and dams used for reproduction and the population
average.
13
Selection Differential (S)
Selection Differential is determine by the
selection intensity applied to the population and
the phenotypic variation present on the
population.
Related to the proportion of the selected animals
to became parents.
Measure of the phenotypic variation on the
population.
14
Selection Differential (S)
S (Sm Sf) / 2

• Superiority is equal to the average of sires and
  dams superiorities used for reproduction.
• In many species, S? gt S?. Why?
• A.I. and E.T. allow to increase S gt A.I. gt E.T.

15
S- Truncation Point
Truncation point
?0
?s
S ?S - ?0 i ? ?p
16
Selection Differential (S)
The smaller is the proportion of animals selected
to be parents bigger is the selection intensity.
In other words, more intensively selection is
been applied. Selection intensity is expressed in
terms of standard deviation from the media.
17
(No Transcript)
18
Superiority is equal to the average of sires and
dams superiorities used for reproduction.
S (Sm Sf) / 2
Selection Intensity is equal to the average of
sires and dams intensities used for reproduction.
i (Im If ) /2
S I average X sp ½ (im if ) sp
19
Selection Differential (S)
20
Standard Deviation
i1 i2 i3
S1gt S2 gt S3!!!!!
21
Selection Differential (S)
The bigger is the selection intensity and
standard deviation bigger is the Selection
Differential.
22
Heritability
The bigger is the Selection Differential and the
Heritability the bigger is the Genetic Gain.
Heritability is proportion of the phenotypic
variance that is Additive (narrow sense) and
consequently inheritable.
The Genetic Gain will be the fraction of the
Superiority of the selected dams and sires that
is inheritable.
!!!
23
Correlation (r)Regression coefficient (b)
Proportion of the phenotypic variation that is
inheritable. h2 can estimated as a regression of
A on P
24
Heritability
The bigger is the heritability the bigger is the
Breeding Value of sires and Dams and consequently
the Genetic Gain.
25
Example 1.
What will be the average progeny phenotype?
26
Example 1.
What will be the average progeny phenotype?
27
Example 1.
What will be the average progeny phenotype?
The average phenotype on the offspring will be
the parents generation average phenotype plus
the genetic gain.
The average phenotype on the offspring will be
73.75 kg.
28
Example 2 ADG in swine

• Initial herd mean 1.7 lb/day (over sexes)
• Selected parent mean 2.0 (over sexes)
• Assume h2 .30

29
Example 2 ADG in swine

• Initial herd mean 1.7 lb/day (over sexes)
• Selected parent mean 2.0 (over sexes)
• Assume h2 .30
• S 2.0 - 1.7 .3 lb/day

30
Example 2 ADG in swine

• Initial herd mean 1.7 lb/day (over sexes)
• Selected parent mean 2.0 (over sexes)
• Assume h2 .30
• S 2.0 - 1.7 .3 lb/day
• ?G h2 ? S (.3)(.3) .09 lb/day

31
Example 2 ADG in swine

• Initial herd mean 1.7 lb/day (over sexes)
• Selected parent mean 2.0 (over sexes)
• Assume h2 .30
• S 2.0 - 1.7 .3 lb/day
• ?G h2 ? S (.3)(.3) .09 lb/day
• ?1 expected herd mean after one generation of
  selection ?0 ?G
• 1.7 .09 1.79 lb/day

32
Example 3 Calf weaning wt (h2 .3)

• Initial herd means Selected parent means
• _________________________________
• 450 females 470 females
• 500 males 600 males
• ______________________________
• 475 avg

33
Example 3 Calf weaning wt (h2 .3)

• Initial herd means Selected parent means
• 450 females 470 females
• 500 males 600 males
• 475 avg

Sfemale 470 - 450 20 lb Smale 600 - 500
100 lb Savg (20 100)/2 60 lb
34
Example 3 Calf weaning wt (h2 .3)

• Initial herd means Selected parent means
• 450 females 470 females
• 500 males 600 males
• 475 avg
• Sfemale 470 - 450 20 lb Smale 600 - 500
  100 lb
• Savg (20 100)/2 60 lb
• R h2 ? S .3(60) 18 lb
• Expected herd mean after one generation of
  selection 475 18 493 lb averaged over sexes

35
Example4 fleece wt (h2.4)

• Initial flock mean 12 lb (over sexes)
• Selected parent means
• 16 lb for males and 14 lb for females (15 over
  sexes)

36
Example 4 fleece wt (h2.4)

• Initial flock mean 12 lb (over sexes)
• Selected parent means
• 16 lb for males and 14 lb for females (15 over
  sexes)
• S 15 - 12 3 lb

37
Example 4 fleece wt (h2.4)

• Initial flock mean 12 lb (over sexes)
• Selected parent means
• 16 lb for males and 14 lb for females (15 over
  sexes)
• S 15 - 12 3 lb
• R h2 ? S .4 (3) 1.2
• Expected flock mean next generation
• 12 1.2 13.2 lb

38
Generation Interval (L)

• L average age of parents when offspring are
  born.
• Cow has calves at ages 2, 3, 4, 5 and 6 yr
• L (2 3 4 5 6)/5 4.0 yr
• Sow give birth at 1, 1.5, 2.25, 2.75 and 3.5 yr.
• L ( 1 1.5 2.25 2.75 3.5)/5 2.2 yr

39
Typical generation intervals (yr)

• cattle
• sheep
• swine
• horses
• chickens

• Males
• 3 - 4
• 2 - 3
• 1.5 - 2
• 8 - 12
• 1 - 1.5

Females 4.5 - 6 4 - 4.5 1.5 - 3 8 - 12 1 - 1.5
40
Use of Gen. Interval (L)

• Calculate selection response per year.
• ?G h2 ? S (per gen.)
• ?G /yr (h2 ? S)/L
• L (L? L?)/2

41
Previous fleece wt example

• ?G h2 ? S 1.2 lb per gen.
• Now, assume L? 3 yr and L? 4.
• L (3 4)/2 3.5 yr

42
Previous fleece wt example

• ?G h2 ? S 1.2 lb per gen.
• Now, assume L? 3 yr and L? 4.
• L (3 4)/2 3.5 yr
• ?G /yr (h2 ? S)/L 1.2/3.5 .343 lb/yr

43
Previous fleece wt example

• ?G h2 ? S 1.2 lb per gen.
• Now, assume L? 3 yr and L? 4.
• L (3 4)/2 3.5 yr
• ?G /yr (h2 ? S)/L 1.2/3.5 .343 lb/yr
• Expected herd mean after 1 yr of selection
• 12 .343 12.343 lb

44
Example ADG in swine (h2.40 and ?P.25 lb/day)

• ?0 1.8 lb/day
• Keep top 60 of females and 20 males.
• i? 1.40 i? .64 iavg 1.02

45
Example ADG in swine (h2.40 and ?P.25 lb/day)

• ?0 1.8 lb/day
• Keep top 60 of females and 20 males.
• i? 1.40 i? .64 iavg 1.02
• R/gen h2 ? i ? ?P .4(1.02)(.25) .102 lb/day
• Expected ?1 ?0 R 1.8 .102 1.902 lb/day

46
Note

• As prop ?, i (and thus S) ?, and so R ?.
• S ? i ? ?P. Thus, S depends on
• 1) p proportion of pop. Selected for breeding
• 2) pop. Variability for trait

47
Compare 3 situations(select for ? YW in beef
cattle)

• a) top 50 selected ?P 100 lb, S 80 lb
• b) top 20 selected ?P 100 lb, S 140 lb
• c) top 20 selected ?P 50 lb, S 70 lb

48
a) top 50 selected ?P 100 lb, S 80 lb
?0
?1
49
b) top 20 selected ?P 100 lb, S 140 lb
?0
?1
50
c) top 20 selected ?P 50 lb, S 70 lb
?1
?0
51
Effectiveness of selection depends on

• 1. Heritability and accuracy of selection (h2)
• 2. Intensity of selection (p, i)
• 3. Population variation (?P, ?A)
• 4. Generation interval (L)
• Note conflict between need for uniformity of
  product versus variability for selection.

52
Alternative pop. situations

• 1) Large VP small VA.
• Implications
• VE large
• h2 low ACC low
• R will be small (selection ineffective)

53
Alternative pop. situations

• 2) Small VE large VA.
• Implications
• h2 and ACC high
• VP moderate to high
• ?G or R will be large (selection very effective)

54
Alternative pop. situations

• 3) Large VE large VA.
• Implications
• VP very large
• h2 could range quite a bit
• Selection progress will be slow, but cumulative
  change over time could be large.

55
Alternative pop. situations

• 4) Small VE small VA.
• Implications
• h2 ACC low to moderate
• Long-term cumulative change will be limited, but
  should get there relatively quickly.

56
Example 1.
What will be the average progeny phenotype?
The average phenotype on the offspring will be
73.75 kg.
57
Note that not all progeny will be 3.75Kg.This
is the average we expect for a large group of
progeny.
58
Within Family VariationWhy do progeny of the
same parents differ?

• Genetic variation within families
• -each individual receive a random one-half of
  genetic material from each parent
• -individuals are not identical to full sibs
• Environmental variation
• -systematic or random environmental variation

59
Adjustment of Performance Records

• Use of Regressions to Adjust Performance Records
• Reduces bias due Environment
• - Increases h2 and thus Accuracy of Selection
• Contemporary Groups

60
Adjustment of Performance Records

• Phenotypes are corrected for known fixed effects
• Fixed effects that can influence phenotype
  include
• -Sex
• -Born as single versus twin
• -Seasonal differences
• For fair comparison phenotypes are adjusted for
  these effects

61
Adjustment of Performance Records
1-Express S as deviation from contemporary group
mean.
Example A 57 Kg male twin has S 57 55Kg 2Kg
62
2- Adjust for effects separately only
applicable in absence of interaction between
effects
Example A 57 Kg male twin is advantaged by 5Kg
for being male and disadvantaged by 2.5Kg for
being a twin. Adjusted S 57-52.5 54.5Kg S (as
deviation from overall mean) 54.5 52.5 2Kg
63
(No Transcript)
64
(No Transcript)
65
CowReplacementRule