sfa3 status on gensap for assessment and optimisation · sfa3 status on gensap for assessment and...

CENTER FOR QUANTITATIVE

GENETICS AND GENOMICS Q AARHUS

UNIVERSITY



UNIVERSITY

SFA3 Status on GenSAP for

assessment and optimisation



UNIVERSITY

Huiming Liu, Tobias Okeno, Kristian Meier, Mark Henryon, Jørn Thomasen, Hadi Esfandyari, Christian Sørensen



UNIVERSITY



UNIVERSITY

What GenSAP can do for us

• Time for development of tools

• Time for uncovering and understanding mechanism underlying consequences of selection

• Both of these support the industry projects



UNIVERSITY



UNIVERSITY

Diminishing marginal returns from

genomic selection as more selection

candidates are phenotyped

T. O. Okeno, M. Henryon and A. C. Sørensen

05-01-2015 3



UNIVERSITY



UNIVERSITY

Hypotheses There is diminishing marginal return from genomic selection as

more candidates are phenotyped

Phenotyping candidates based on a priori information is beneficial

There is a best distribution of phenotypes on the sexes with respect to genetic gain and inbreeding rate

05-01-2015 4



UNIVERSITY



UNIVERSITY

05-01-2015 5

Genetic gain

0,85

0,87

0,89

0,91

0,93

0,95

0,97

0,99

1,01

1,03

1,05

20 40 60 80 100Phenotyping proportion (%)

ΔG, genotypic standard deviation

Estimated breeding value

Random selection



UNIVERSITY



UNIVERSITY

6 05-01-2015

Genetic gain

0,85

0,87

0,89

0,91

0,93

0,95

0,97

0,99

1,01

1,03

1,05

20 40 60 80 100

Phenotyping proportion (%)

ΔG, genotypic standard deviation

Estimated breeding value

Random selection

Sex differentiation

No sex differentiation



UNIVERSITY



UNIVERSITY

05-01-2015 7

0,84

0,86

0,88

0,90

0,92

0,94

0,96

0,98

1,00

100:0 75:25 50:50 25:75 0:100

Male:female phenotyping ratio

ΔG, phenotypic standard deviation

50%

40%

30%

20%

Genetic gain for Male:Female ratio



UNIVERSITY



UNIVERSITY

Diminishing return as more candidates are phenotyped

Use of a priori information to select animals to phenotype is beneficial

Mainly phenotyping sex with high selection intensity is beneficial at low

phenotyping proportions

Less intensively selected sex should also be considered at high

phenotyping proportions

05-01-2015 8

Conclusions



UNIVERSITY



UNIVERSITY

Specific challenges adressed in GenSAP SFA3

• Extended tool box

• Non-additive genetic variation (Hadi)

• Selection across multiple environments

• Use of genomic information for handling inbreeding (Huiming)

• Proof-of-concept (Kristian)



UNIVERSITY



UNIVERSITY

Maximizing crossbred performance through purebred genomic selection

Hadi Esfandyari

Christian Sørensen

Piter Bijma



UNIVERSITY



UNIVERSITY

Aims

• To investigate the benefits of GS of purebreds for CP based on purebred information

• To compare separate pure line reference populations with a joint reference population



UNIVERSITY



UNIVERSITY

Predicting crossbred performance

• Predicting additive and dominance effects of each marker

• Purebred breeding value of an animal depends on allele frequency in its own breed

• Crossbred breeding value of an animal depends on allele frequency in the other breed



UNIVERSITY



UNIVERSITY

Performance of crossbreds



UNIVERSITY



UNIVERSITY

Conclusions

• Selecting the pure lines for crossbred performance increases phenotypic performance of crossbreds

• Selecting for crossbred performance increases the expression of heterosis

• If the breeds are closely related a joint reference population is beneficial



UNIVERSITY



UNIVERSITY


GENETICS AND GENOMICS Q

Huiming Liu

Maintenance of genetic diversity in genomic selection

AARHUS

UNIVERSITY

15

Supervisors: Elise Norberg Peer Berg Christian Sørensen Theo Meuwissen



UNIVERSITY



UNIVERSITY

How to maintain the genetic diversity?

16

0.0 0.2 0.4 0.6 0.8 1.0

24

68

10

Low allele freq. ~ High allele freq.

We

igh

t (W

)

• Putting more weight on the rare favourable alleles (WGEBV)

0.00 0.01 0.02 0.03 0.04 0.05 0.06

0.1

10.1

20.1

30.1

40.1

50.1

60.1

7

Fped

genetic g

ain

-1

-5

-10

-25

-50

Ge

net

ic g

ain

• Optimum contribution selection (OCS)

Rate of inbreeding



UNIVERSITY



UNIVERSITY

Selection design

17

rule

criterion

Truncation

selection

Optimum Contribution

Selection

A G

GEBV

GEBV

OCSA

OCSG

WGEBV

WGEBV

WOCSA

WOCSG



UNIVERSITY



UNIVERSITY

Genetic gain

18

rule

criterion

Truncation

selection

Optimum Contribution

Selection

A G

GEBV

GEBV

0.435

OCSA

0.542

OCSG

0.587

WGEBV

WGEBV

0.495

WOCSA

0.590

WOCSG

0.604

13



UNIVERSITY



UNIVERSITY

Conclusion

19

The strategy that combines WAF and OCS is very promising



UNIVERSITY



UNIVERSITY

Genomic selection in mink (Neovison vison): A simulation study



UNIVERSITY

Kristian Meier Anders Christian Sørensen

Janne P. Thirstrup Mogens Sandø Lund



UNIVERSITY



UNIVERSITY

Breeding plan without genomic information



UNIVERSITY

240 males 1200 females

6000 offspring

40% reused second year

Breeding value estimation

Selection

Weight, feed efficiency, pelt quality

(live animals)

Skin quality

Litter size, barren females

Pelting

Proces starts over

Challenge: No registration on selection candidates for: - Litter size, - Barren females - Skin quality



UNIVERSITY



UNIVERSITY

• Do we need to make DNA analysis on all selection candidates?

Simulation design

• How high accuracy do we get with genomic selection in mink? – We don´t know?

– We simulate scenarios with low, medium and high accuracy



UNIVERSITY

Number of DNA analysis

Selection candidates Number

All 6000

Best 10% males 300



UNIVERSITY



UNIVERSITY



UNIVERSITY

Genomic selection

Total economic gain Dkr pr.

female pr. year Accuracy DNA analysis

High All 97

10 % best males 71

Low All 60

10 % best males 57

Breeding plan without genomic information 54

Monetary genetic gain



UNIVERSITY



UNIVERSITY

Genetic gain in traits



UNIVERSITY

Genomic selection

Contribution (%) from five traits to the total economic gain

Litter size

Weight

Barren females

Pelt quality

Feed efficiency

Accuracy DNA analysis

High All 42 19 19 0 20

10 % best males

26 37 12 -5 30

Low

All 12 51 7 -8 39

10 % best males 10 54 4 -9 41

Breeding plan without genomic information 4 59 2 -10 45



UNIVERSITY



UNIVERSITY

Conclusion



UNIVERSITY

• Genomic selection increases total economic gain

• Increased genetic gain for litter size, barren females and pelt quality

• Even at low accuracy and few DNA analyses genetic gain is increased



UNIVERSITY



UNIVERSITY

Future work



UNIVERSITY

• Evaluate different economic weights

• Cost (DNA) benefit (economic gain) analysis

• Future infrastructure supporting genomic selection



UNIVERSITY



UNIVERSITY

Wrap-up

• So far, we have just scratched the surface

• Recruitment allows us to speed up from January – PhD at NMBU started September 1st 2014

– Post Doc jointly with SFA1 starting January 1st 2015

– Post Doc assisting with ADAM programming starting January 1st 2015



UNIVERSITY



UNIVERSITY

Specific challenges adressed in GenSAP SFA3

• Extended tool box (Huiming, Jørn, Mark, Beatriz)

• Non-additive genetic variation (Hadi)

• Selection across multiple environments

• Use of genomic information for handling inbreeding (Huiming, Gebreyohans, Tobias)

• Proof-of-concept (Kristian)

sfa3 status on gensap for assessment and optimisation · sfa3 status on gensap for assessment and...

Documents