Making the Web equal Profit Surfing for Genetics

1
Making the Web equal ProfitSurfing for Genetics

• Dorian Garrick Mark Enns
• Department of Animal Sciences
• Colorado State University

2
Surfing for Genetics

• Surfing for Convenience
• Surfing to Support Decisions based on your own
  Customized Computations

3
Convenience

• Finding a Particular Bull/Breed/Breeder
• Sort Orders
• Finding extreme bulls for some attribute
• Filters
• Finding bulls with particular combinations of
  attributes
• Up-to-date EPD and ACC information

4
Customized Computations

• Interpretation of Threshold Traits
• Interactions between ERTs
• Assessment of Nutritional Implications
• Assessment of Financial Implications
• Perhaps also Economic Implications
• Accounting for Risk
• Multibreed Evaluation Crossbreeding

5
Interpretation of a Typical EPD
W W D 50 lb
W W D 20 lb
6
Interpretation of a Typical EPD
W W D 50 lb
W W D 20 lb
Herd 1 Average 500 lb Average 530
7
Interpretation of a Typical EPD
W W D 50 lb
W W D 20 lb
Herd 1 Average 500 lb Average 530 Herd 2 Average
550 lb Average 580
8
Interpretation of Threshold Traits

• Calving Ease EPD
• Stayability EDP
• Heifer Pregnancy EPD

9
Underlying Scores to Preg Rate
Difficult to get pregnant
Easy to get pregnant
Average
10
Underlying Scores to Preg Rate
Difficult to get pregnant
Easy to get pregnant
Average
Suppose 20 heifers are open And 80 heifers are
pregnant
11
Underlying Scores to Preg Rate
Pregnant Heifers
Heifers not in calf
20
Difficult to get pregnant
Easy to get pregnant
Average
12
Underlying Scores to Preg Rate
Pregnant Heifers
Heifers not in calf
20
Truncn pt 0.84s
Threshold
13
Underlying Scores to Preg Rate
Pregnant Heifers
Heifers not in calf
20
Truncn pt 0.84s
0.38
14
Underlying Scores to Preg Rate
Pregnant Heifers
Heifers not in calf
20
Truncn pt 0.84s
12
Truncn pt 0.84 0.38/1.171.165
Phenotypic s.d. 1.17
0.38
15
Underlying Scores to Preg Rate
10
Truncn pt 1.28s
5.5
Truncn pt 1.28 0.38/1.171.605
Phenotypic s.d. 1.17
0.38
16
Sensitive to the Average

• An underlying EPD of 0.38 for heifer pregnancy
  would increase pregnancy rate
• By 8.0 if average pregnancy rate is 80
• By 4.5 if the average is 90
• Phenotypic interpretation of underlying
  threshold scores depends upon the mean
• Published values are at a mean of 50

17
Solution

• Publish values that are hard to interpret
• OR Publish tables of EPDs relevant to different
  average levels of performance
• Calving Ease
• First Calf 75, 80, 85, 90, 95
• Mixed Age 95, 99
• Stayability 40, 45, 50, 55, 60
• Heifer Pregnancy 75, 80, 85, 90, 95

18
Solutions (cont)

• OR Use web-based decision support
• User defined average levels of performance
• Compute the
• number of pregnant vs open heifers
• number of easy vs difficult calvings
• Likely age structure of the herd
• Number of replacement heifers required
• Number of cull yearlings and mixed age cows

19
Suppose our goal is incr sale wt
W W D 20 lb
W W D 50 lb
Base30 lb
Base
20
Interactions between ERTs

• WWD EPD 30 lb (all other EPDs equal)
• gives 30 lb weanlings

21
Interactions between ERTs

• WWD EPD 30 lb (all other EPDs equal)
• gives 30 lb weanlings
• gives 22 lb weanling sale wt per cow in a
  typical self-replacing herd

22
Interactions between ERTs

• WWD EPD 30 lb (all other EPDs equal)
• gives 30 lb weanlings
• gives 22 lb weanling sale wt per cow in a
  typical self-replacing herd
• Increasing ST 8 will give another 8 lb

23
Interactions between ERTs

• WWD EPD 30 lb (all other EPDs equal)
• gives 30 lb weanlings
• gives 22 lb weanling sale wt per cow in a
  typical self-replacing herd
• Increasing ST 8 will give another 8 lb
• Increasing HPG 12 will give another 3lb

24
Interactions between ERTs

• WWD EPD 30 lb (all other EPDs equal)
• gives 30 lb weanlings
• gives 22 lb weanling sale wt per cow in a
  typical self-replacing herd
• Increasing ST 8 will give another 8 lb
• Increasing HPG 12 will give another 3lb
• Increasing CED 11 will give another 1 lb

25
Interactions between ERTs

• Many ERT interact in a system context
• Total sale weight at weaning is altered by WWD,
  WWM, STAY, HPG, CED, CEM, ME (plus BW YWT)
• The impact of any one EPD on the change in sale
  weight depends upon all the other EPDs and the
  average levels of performance

26
Assessment of Nutritional ( other input)
Implications

• Feed requirements are influenced by the
• Expected maintenance requirements
• Expected requirements for gain
• Deviation from our expectations (known as
  residual feed intake or RFI)
• Changing any of WWD, WWM, STAY, HPG, CED, CEM,
  ME, BW, YW will alter whole herd feed requirements

27
Assessment of Financial Implications

• Changes in profit can be calculated from change
  in income prices change in expenses costs
• Straightforward (but tedious) arithmetic
• Prices Costs can be obtained on a financial
  basis or an economic basis
• What is the cost of feed in an extensive cow-calf
  grazing operation?

28
Economic Cost of Feed

• Feed cost can be calculated from its
  opportunity cost
• Measure of what net income would be given up if
  you had less feed (and less cows)
• Can be calculated from comparing the system
  profit of two herds of alternative genotypes with
  different stocking rates such that each consume
  the same amount of feed

29
Solutions

• Leave it to bull buyers to struggle thru facts
  calculations
• OR put it all together via the web
• Website ert.agsci.colostate.edu

30
(No Transcript)
31
(No Transcript)
32
Current Philosophical Approach
Base Situation
Perturbed Situation
Current (equilibrium) Cow Herd (EPD Performance)
Like Merit Bulls
X
Base Cow-calf outputs inputs
33
(No Transcript)
34
(No Transcript)
35
Current Philosophical Approach
Base Situation
Perturbed Situation
Current Cow Herd (EPD)
Chosen Bulls
X
Current (equilibrium) Cow Herd (EPD Performance)
Like Merit Bulls
Daughter (equilibrium) Cow Herd (EPD Base
mean Performance)
X
Base Cow-calf outputs inputs
36
Current Philosophical Approach
Base Situation
Perturbed Situation
Current Cow Herd (EPD)
Chosen Bulls
X
Current (equilibrium) Cow Herd (EPD Performance)
Like Merit Bulls
Daughter (equilibrium) Cow Herd (EPD base
mean Performance)
Like Merit Bulls
X
X
Perturbed Cow-calf outputs inputs
Base Cow-calf outputs inputs
37
(No Transcript)
38
Accounting for Risk

• Consider the following three bulls
• Bull prog Acc Profit
• Lima 30 0.5 to 0.6 908
• Sierra 240 0.7 to 0.8 729
• Bravo 3700 0.8 to 0.95 648

39
Accounting for Risk

• On average, true EPD is equally likely to be
  greater or lesser than published
• ACC allows us to quantify the extent to which the
  estimate may vary from true
• Considering just BW, WW, YW, Milk, ME and not (in
  this example) CED, CETM, HPG, ST we can compute
  many possible realizations of each bull

40
Extra Profit per 30 daughter-years
3700 progeny
648
41
Accounting for Risk
3700 progeny
240 progeny
729
42
Accounting for Risk
3700 progeny
240 progeny
30 progeny
908
43
Solution

• Publish an expected change table
• OR provide web options for quantifying risk
  (prototype available this Fall)

44
Multibreed Evaluation Xbreeding
A1 S3 are breed average EPDs
45
Multibreed Evaluation Xbreeding
Angus Base
A1 S3 are breed average EPDs
46
Multibreed Evaluation Xbreeding
47
Multibreed Evaluation Xbreeding
48
Multibreed Evaluation Xbreeding
49
Multibreed Evaluation Xbreeding
50
Multibreed Evaluation Xbreeding
51
Multibreed Evaluation Xbreeding
52
Multibreed Evaluation Xbreeding
53
Solution

• Publish within-breed EPDs
• Let users find breed adjustments heterosis
• Publish multibreed EPDs
• Let users deal with heterosis coefficients
  heterosis
• Publish all EPDs on a multibreed base with
  heterosis factors included according to the breed
  of dam for every ERT
• Use web-based decision support

54
Summary

• To date, the major value of the web has been
  convenience
• In future, the web will provide an interface to
  knowledge (eg nutritional requirements and
  heterosis factors) and information that for
  customized calculations to support your decisions
• Better decision support will give better
  decisions (eg more profit)