Inbreeding Using Genomics and How it Can Help. Dr. Flavio S. Schenkel CGIL- University of Guelph

Transcription

1 Inbreeding Using Genomics and How it Can Help Dr. Flavio S. Schenkel CGIL- University of Guelph

2 Introduction Why is inbreeding a concern? The biological risks of inbreeding: Inbreeding depression Accumulation of deleterious alleles Reduction of genetic variance Loss of genetic diversity

3 Inbreeding depression Trait Loss per 1% inbreeding Fat yield (kg) 1.1 Protein yield (kg) 0.5 Conformation (points) 0 Days open (days) 1.4 Calf survival 1 st calving (%) 0.5 Productive life (days) 13 (Van Doormaal, CDN report. March 2008)

4 Genetic variance Loss of genetic variance: e.g. Canadian Holstein 460, ,000 Additive genetic variance 450,000 Milk yield [kg 2 ] 445, , , , , , , , Year of birth (Stachowicz Ph.D. Thesis, UofG)

5 Genetic variance Loss of genetic variance: Genomic selection Change in genetic variance (%) after 10 generations Number of genes Selection scheme Few Many Genomic Selection Traditional Selection No selection (random) (Adapted from Bastiaansen et al Gen. Sel. Evol. 44:3)

6 Genetic diversity Loss (%) of genetic diversity: e.g. Canadian Holstein % Loss Year (Stachowicz et al JDS 94: )

7 Introduction Inbreeding has been a concern for long time. Would genomic selection change this? No, it will be an even bigger concern. - Faster genetic progress associated with: Shorter generation interval Higher inbreeding rate per year No time for selection to counter balance negative effects of inbreeding

8 Faster genetic progress An illustration (Van Doormaal, CDN report. July 2012)

9 Consequences of genomic selection Selection Scheme Inbreeding per generation (%) Generation interval (years) Inbreeding per year (%) Genetic change per generation (%) Conventional Genomic: Turbo (Adapted from Buch et al J. Anim. Breed. Gen. 129: )

10 Consequences of genomic selection At genomic level IBD_T F AVE IBD_T AVE IBD_Q QTL EFF P DIFF E+00 IBD / Inbreeding E E E E-01 QTL effect / Difference in allele frequency Genome 0.00E+00 IBD_T= True IBD, F= Pedigree-based Inbreeding, AVE IBD_T= Average IBD_T, AVE IBD_Q= Average IBD at the QTLs, QTL EFF= QTL effect, P DIFF= Change in QTL allele frequency (Stachowicz, Ph.D. Thesis, UofG)

11 Average Inbreeding - All animals Introduction Is the observed rate of inbreeding going up with genomic selection? Yes, there could be some evidence already. E.g. Canadian Holstein Inbreeding coefficient Year of birth Inbreeding coefficient Average Inbreeding - All animals Year of birth

12 Introduction Would an extra attention on inbreeding with genomics be needed? Yes - Breeding age of daughters of a bull and availability of semen from his sons are aligned. Risk of mating of siblings - Shorter generation interval, faster inbreeding rates. Co-selection of long haplotypes Higher accumulation of deleterious recessives

13 Co-selection of haplotypes Traditional Pedigree Sire of Sire Animal 1 Animal 2 Sire Dam Dam of Sire Sire of Dam Dam of Dam

14 Co-selection of haplotypes Genomic Pedigree ctgtagcgatcg atgtcgctcacg ctgtctagatcg atggatcgatcg ctgtagcgatcg cgatctagatcc agagatcgatcg atgtcgctcacg atagatcgatcg ctgtagcttagg agggcgcgcagt cgatctagatcc cggtagatcagt agagatcgatcg atggcgcgaacg ctatcgctcagg Higher selection pressure in parts of the genome (haplotypes)

15 Co-selection of haplotypes Measurement of inbreeding by chromosome segments instead of individual locus (From Pryce et al JDS 95 : )

16 Co-selection of haplotypes Individuals with similar genomic or pedigree relationships can have quite different proportions of chromosome segments that are identical by descent (IBD). If deleterious homozygotes are more likely to arise as a consequence of recent inbreeding, then strategies to minimize IBD chromosome segments could be a way of reducing them (Pryce et al. 2012).

17 Possibilities with genomics Assessment of inbreeding at genomic level - Under genomic selection, pedigree-based inbreeding does not reflect well the true genomic inbreeding IBD_T F AVE IBD_T AVE IBD_Q 0.2 IBD / Inbreeding Genome

18 Possibilities with genomics - Under genomic selection controlling inbreeding using pedigree or genomic relationships has different effectiveness on genomic inbreeding. Target Genetic Pedigree Genomic Inbreeding Change Inbreeding Inbreeding Constraint on pedigree inbreeding Constraint on genomic inbreeding (Adapted from Sonesson et al th WCGALP)

19 Possibilities with genomics Controlling inbreeding using pedigree (P) or genomic (G) relationships Impact on the genome = Using P = Using G = Variance (From Sonesson et al th WCGALP)

20 Selection and Mating Systems Selection system: Truncated selection: TS (select animals with top evaluations as parents) Optimum selection: OS (selection based on both genetic progress and inbreeding) Mating system: Minimum Co-ancestry: MC (mate animals that are least related as possible) Random Mating: RM (mate animals at random)

21 Selection and Mating Systems Optimum selection (OS) vs. Truncated Selection (TS) Minimum Co-ancestry (MC) vs. Random mating (RM) Traditional evaluation TS & RM TS & MC OS & RM OS & MC TR EBV RM TR EBV MC OC EBV RM OC EBV MC h 2 = 0.30 Inbreeding Generation (Adapted from Stachowicz, Ph.D. thesis, UofG)

22 Selection and Mating Systems Optimum selection (OS) vs. Truncated Selection (TS) Minimum Co-ancestry (MC) vs. Random mating (RM) Genomic evaluation TS & RM TS & MC OS & RM OS & MC TR GAS RM TR GAS MC OC GAS RM OC GAS MC h 2 = 0.30 Inbreeding Generation (Adapted from Stachowicz, Ph.D. thesis, UofG)

23 Selection and Mating Systems Genome-wide true inbreeding after 30 generations 0.25 TR EBV Traditional RM Evaluation TR EBV MC TR GAS RM Genomic Evaluation TR GAS MC TS & RM TS & MC TS & RM TS & MC h 2 = IBD Genome (Adapted from Stachowicz, Ph.D. thesis, UofG)

24 Selection and Mating Systems Phenotypic gain after 30 generations TR TS h^2=0.30 OC OS h^2= Phenotypic values EBV RM RM EBV MC MC GAS RM RM GAS MC MC Traditional Evaluation Genomic Evaluation (Adapted from Stachowicz, Ph.D. thesis, UofG)

25 Possibilities with genomics Discovery and test for unfavorable or deleterious recessive alleles/haplotypes: Examples: - E.g. CVM (Thomsen et al. 2006) - Fertility haplotypes (VanRaden et al. 2011)

26 Possibilities with genomics Haplotypes impacting fertility in Holstein Impact on Haplotype Freq CR NRR Earliest Known Ancestor(s) HH1 4.5% -3.1% -1.1% Pawnee Farm Arlinda Chief HH2 4.6% -3.0% -1.7% Willowholme Mark Anthony HH3 4.7% -3.2% -3.1% Gray View Skyliner Glendell Arlinda Chief (Adapted from VanRaden et al JDS 94 : )

27 Conclusions The availability of high-density SNP data has revolutionized dairy cattle breeding, changing breeding schemes and leading to higher rates of genetic gain. Higher rates of genetic gain from genomic selection will be associated with much higher annual rates of inbreeding, so methods to control inbreeding will become increasingly important.

28 Conclusions For traditional selection, pedigree-based inbreeding is an appropriate measure of true inbreeding across the genome. However, for genomic selection, inbreeding around selected markers can be much higher than the estimated pedigree-based inbreeding. This requires re-evaluate what is an appropriate rate of inbreeding when directly measured in the genome.

29 Conclusions Continued use of mating strategies, such minimum co-ancestry mating, is even more important under genomic selection. The application of optimum contribution selection could further help in controlling inbreeding. The use of genomic relationship information could better control inbreeding surrounding the markers.

30 Conclusions Genomic information can provide new tools for monitoring genome-wide inbreeding and controlling the frequency of deleterious recessive alleles/haplotypes.

31 Acknowledgments Ph.D. student: Katarzyna Stachowicz Pedigree data from: The Canadian Dairy Network (CDN) Financial support from: DairyGen Council of CDN Natural Sciences and Engineering Research Council of Canada (NSERC)

32 Thank you