potential and pitfalls for genomic selection- chad dechow

39
Potential and Pitfalls for Genomic Selection

Upload: dairexnet

Post on 10-May-2015

1.194 views

Category:

Education


1 download

DESCRIPTION

Dr. Chad Dechow presented this information for DAIReXNET on Monday, January 14, 2013. For more information, please see our archived webinars page at www.extension.org/pages/15830/archived-dairy-cattle-webinars.

TRANSCRIPT

Page 1: Potential and Pitfalls for Genomic Selection- Chad Dechow

Potential and Pitfalls for Genomic Selection

Page 2: Potential and Pitfalls for Genomic Selection- Chad Dechow

Topics

• Review of genomic technology and implementation 4-path model

• Comparisons of early genomic predictions to actual daughter proofs Traits to be careful Who should be using genomics, who not? Spread risk

• Genomics as a herd management tool• Inbreeding• Beyond SNPs

Page 3: Potential and Pitfalls for Genomic Selection- Chad Dechow

From Phenotype to Genotype:diacylglycerol acyltransferase 1

• Enzyme involved in triglyceride synthesis Chromosome 14 Knockout mice: complete absence of milk production

• Bi-nucleotide substitution: lysine to alanine +300 lbs milk +5 lbs protein +.17% fat -13 lbs fat Fatty acid profiles altered

• Terrific – but…Grisart et al., 2002

Page 4: Potential and Pitfalls for Genomic Selection- Chad Dechow

Whole Genome Approach

• Single nucleotide polymorphisms 10 – 50 million present in genome Not inherited independently of each other

• Tests Bovine SNP 50

• Cost: $125 Low density

• 9,000 currently (replaces 6K, which replaced 3K)• Used to “impute” 50K• Cost: $45

High density• ~777,000• Early research has not been exciting• Cost: $250

Page 5: Potential and Pitfalls for Genomic Selection- Chad Dechow
Page 6: Potential and Pitfalls for Genomic Selection- Chad Dechow

Association of SNP with Fat Yield

Page 7: Potential and Pitfalls for Genomic Selection- Chad Dechow

Association of SNP with Final Score

Page 8: Potential and Pitfalls for Genomic Selection- Chad Dechow

Genetic Progress

• How does this speed genetic progress?

IntervalGeneration

ianceGeneticVarntensitySelectionIyreliabilitYearG

**/

Calf

SireSire of Sire

Dam of Sire

DamSire of Dam

Dam of Dam

1.Lower generation interval2.Higher accuracy for females3.Selection Intensity

Page 9: Potential and Pitfalls for Genomic Selection- Chad Dechow

Implementation

• First official proofs in January of 2009• Quickly adopted

Sires of sons – vast majority• Marketing differs by

bull stud Mixed lineup separate lineups

2008 201105

101520253035404550

Young sire matings

Holstein Jersey

Perc

ent

Page 10: Potential and Pitfalls for Genomic Selection- Chad Dechow

Comparison of Jan 2009 to Dec 2012 Daughters Deviations

517 bulls0 daughters in 2009 and ≥100 daughters currently

R² = 0.546563821115401

Milk Yield

2009 PTAM

2012

Dau

Yie

ld D

evia

tion

R² = 0.340733769228037

Productive Life

2009 PTAPL

2012

Dau

Dev

iatio

n

Page 11: Potential and Pitfalls for Genomic Selection- Chad Dechow

Realized Reliabilities

Milk yield Daughter Preg Rate Productive Life0%

10%

20%

30%

40%

50%

60%

70%

80%

HolsteinJerseyBrown Swiss

Page 12: Potential and Pitfalls for Genomic Selection- Chad Dechow

Top 25 Young Sires and Proven Bulls in 2009

Average 2009 Average 2012 Top 20120

100

200

300

400

500

600

700

800

900

Genomic YSProven

Page 13: Potential and Pitfalls for Genomic Selection- Chad Dechow

Net Merit Changes

Aug-08

Nov-08

Feb-09

May

-09

Aug-09

Nov-09

Feb-10

May

-10

Aug-10

Nov-10

Feb-11

May

-11

Aug-11

Nov-11

Feb-12

May

-12

Aug-12

0100200300400500600700800900

1000

FreddieCassinoSholtonAtwood

$

Page 14: Potential and Pitfalls for Genomic Selection- Chad Dechow

Traits to watch

• Productive Life Must wait for cows to die Predictors to help

• Calving related traits

Body Size

Udder Feet & Legs

DPR SCS-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

Productive Life Genetic Corre-lations

Previous Current

Page 15: Potential and Pitfalls for Genomic Selection- Chad Dechow

Who Should Use Genomic Young Sires?

Use• Involved with marketing

Will have hits and misses Goes with the territory

• Not marketing Watching calving traits on

virgin heifers Spreading risk by using a

selection Willing to accept some

misses

Do not use• Not marketing• You want to minimize

calving issues• Willing to miss out on the

best for 3 years Average may not be

different, but top will be lower

Page 16: Potential and Pitfalls for Genomic Selection- Chad Dechow

Beyond Sire Selection

Page 17: Potential and Pitfalls for Genomic Selection- Chad Dechow

DNA Level Mating Decisions

• Replacement for visual appraisal mating programs?

• Chromosome level mating http://

aipl.arsusda.gov/CF-queries/Bull_Chromosomal_EBV/bull_chromosomal_ebv.cfm

Use 17 digit ID style (HOUSA000000000000) Cows entered on same page as bulls

Page 18: Potential and Pitfalls for Genomic Selection- Chad Dechow
Page 19: Potential and Pitfalls for Genomic Selection- Chad Dechow

Can we Improve Her?

23 gallons/day for a year

Page 20: Potential and Pitfalls for Genomic Selection- Chad Dechow

Haplotype Projections: Milk

Brown Swiss Holstein Jersey0

10000

20000

30000

40000

50000

60000

70000

80000

90000

Largest DGV Lower Bound Upper Bound

Sele

ction

Lim

it M

ilk (l

bs)

Cole et al., 2011

Page 21: Potential and Pitfalls for Genomic Selection- Chad Dechow

Haplotype Projections: DPR

Brown Swiss Holstein Jersey0

20

40

60

80

100

120

140

160

Largest DGV Lower Bound Upper Bound

Sele

ction

Lim

it D

PR

Cole et al., 2011

Page 22: Potential and Pitfalls for Genomic Selection- Chad Dechow

Opportunity 2013

• Only bull studs can genotype males 6 Studs• Contributed $ and DNA

License agreement• Newer chips detect Y chromosome genes• Agreement ends in 2013• If you have a good bull, do you sell him?

Market your own bull? What will it cost?

Page 23: Potential and Pitfalls for Genomic Selection- Chad Dechow

Genomics as a Herd Management Tool

• Premise: Genomics can play a role for commercial milk producers with excess heifers

• Helpful link http://edis.ifas.ufl.edu/pdffiles/AN/AN27000.pdf

Page 24: Potential and Pitfalls for Genomic Selection- Chad Dechow

NY-PA Replacement Rates

Page 25: Potential and Pitfalls for Genomic Selection- Chad Dechow

NY-PA Cull Rates

Page 26: Potential and Pitfalls for Genomic Selection- Chad Dechow

Maintaining Herd Size

• More replacements than needed Increase cull rate?• Fewer problem cows• Less “mature milk”

Sell heifers?• Lower feed costs• Heifer market sustainable?

Page 27: Potential and Pitfalls for Genomic Selection- Chad Dechow

Selling Heifers

• Value of testing• Herd improvement by culling the bottom end

70%, 80%, or 90% of calves kept What happens to the value of my remaining

calves if I genomically test first? What is the $ Net Present Value of testing?

**First culling threshold: sick/diseased calves

Page 28: Potential and Pitfalls for Genomic Selection- Chad Dechow

$Net Merit of Remaining Calves

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0

50

100

150

200

250

90% kept 80% kept 70% kept

% of calves tested

$ N

et M

erit

Page 29: Potential and Pitfalls for Genomic Selection- Chad Dechow

$Value = $NM – Test Cost

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%0

20

40

60

80

100

120

140

90% Kept 80% Kept 70% Kept

% Tested

$ N

et M

erit

Page 30: Potential and Pitfalls for Genomic Selection- Chad Dechow

Net Present Value

• We don’t need to test every calf Top sires will rarely have offspring you want to

cull• Net Present Value compared with parent

average selection

Page 31: Potential and Pitfalls for Genomic Selection- Chad Dechow

What to Sell

• Lots of heifers = limited marketing potential Save on feed costs

• Beef sires Male sexed semen Gaining traction Helpful with Jerseys

Page 32: Potential and Pitfalls for Genomic Selection- Chad Dechow

Individualized Cow Management?

• Should we alter management to accommodate genetic potential? High dairy form = high early lactation BCS loss

risk• Calving BCS should be LOW

Lower yield potential• Breed back more quickly?

• Group cows by genetic potential?

Page 33: Potential and Pitfalls for Genomic Selection- Chad Dechow

Will Genomics Impact Inbreeding Rates?

Page 34: Potential and Pitfalls for Genomic Selection- Chad Dechow

Close Inbreeding (F=14.7%): Double Grandson of Aerostar

Aerostar

Aerostar

Megastar

Chromosome 24

Megabuck

Digne

VanRaden, 2008

Page 35: Potential and Pitfalls for Genomic Selection- Chad Dechow

• Likely to accelerate with genomics Shorter generation interval Technology is “pattern recognition”• Unusual genetic make-up = unrecognized pattern

• Line developmentAerostar

AerostarMegastar

Chromosome 24

Megabuck

Digne

Inbreeding

Identical by descent = inbred

Page 36: Potential and Pitfalls for Genomic Selection- Chad Dechow

If we know the DNA code

• Why are genomic tests 100% accurate? Markers are random & may have nothing to do

with performance themselves Copy number variation Not accounting for dominance/gene interactions “Epigenetic” effects• Alter gene expression independently of DNA code• High milk yield during gestation = lower milk yield

daughter?

Page 37: Potential and Pitfalls for Genomic Selection- Chad Dechow

The more we learn, the less we know• Intelligent design cannot explain the presence of a

nonfunctional pseudogene … the designer made serious errors, wasting millions of bases of DNA … junk … Evolution, however, can explain them easily … they persist in the genome as evolutionary remnants of the past history (Miller, 1994)

Page 38: Potential and Pitfalls for Genomic Selection- Chad Dechow

Marker Effects

Page 39: Potential and Pitfalls for Genomic Selection- Chad Dechow

Thank you and are there any questions?