PresentationPDF Available

Utilisation of the ovine HD SNP chip for meat quality traits in NZ composite terminal sire breeds

Authors:

Abstract

FarmIQ (www.farmiq.co.nz) a joint New Zealand government and industry Primary Growth Partnership programme, in conjunction with Illumina and the International Sheep Genomics Consortium (ISGC) released the Ovine Infinium®HD SNP BeadChip in 2013. FarmIQ commissioned the HD Chip as part of its mission to add value to red meat by improving linkages between animals’ meat yield and quality, and what happens on the farm and in the processing plant. The ISGC together with FarmIQ developed the HD SNP chip following the re-sequencing of 73 diverse breeds from around the world and utilised the recently released sheep genome assembly (OARv3.1). The chip contains ~600 000 SNPs; although majority (~90%) of SNPs have MAF>0.10, there are also rare genetic variants and ~53,000 SNPs that are known to affect proteins included. The FarmIQ programme has now used this HD chip for genome wide association studies and multi-breed genomic predictions for meat quality traits in New Zealand composite terminal sire breeds. Preliminary results suggest genomic prediction accuracies will be similar to measuring the trait on the animal itself and most meat quality traits also have several loci of large effect. https://pag.confex.com/pag/xxiii/webprogram/Paper15089.html
Utilisation of the ovine HD SNP chip
for meat quality traits in NZ
composite terminal sire breeds
Shannon Clarke, Michael Lee, Rudiger Brauning, Wendy Bain, Tracey Van Stijn,
Rayna Anderson, Ken Dodds, John McEwan
& the Animal Genomics Team
www.farmiq.co.nz
a joint New Zealand government and industry Primary Growth
Partnership programme
Bringing farmers and consumers together
Farmers understand how to produce animals that meet
consumer preferences and receive payments based on
meat quality.
Consumers are offered premium-branded red meat that
consistently meets their eating quality preferences.
Genetics programme in FarmIQ
deliver high-quality genetic and genomic breeding value predictions
growth, meat yield and eating quality in terminal sire flocks
no real selection for meat quality in New Zealand sheep previously
Lamb is a premium-priced product and should be a good eating experience
Develop tools to tailor products to consumer specifications
Concentrated on terminal sire breeds (~80 percent of the leading industry
sires)
Deliver increased profitability for each of the value chain participants
(farmer, processor, retailer)
Four phases
Phase 1: HD SNP chip development
Phase 2: Identify and measure eating quality and
carcass yield traits in progeny of terminal sires.
Phase 3: Genotyping individuals and prediction
equation development
Phase 4: Beta test results in commercial terminal sire
breeding flocks to ensure the system is accurate,
practical and cost-effective
1. Genomic Selection
- increased marker density ensures better marker trait association
- increased prospect for successful across breed prediction
- part of tool kit (+ GWS + SNP50 + 5K + imputation) to drive
genotyping efficiency
2. Broader Experimental Possibilities
- Loss of function SNPs
- facilitates merging with gene expression studies: eQTL
- rare SNP
Phase 1: HD SNP Chip development
HD chip development
Baylor CSIRO DEPI AgR
BAM
SNPs & INDELs
(VCF)
ATLAS
Mappings
(SNPs, refseqs,…)
SNPs & INDELs
(VCF)
SAMTOOLS
OARv3.1 OARv3.1 OARv3.1 OARv3.1
BAM BAM BAM
SNP filtering SNP filtering SNP filtering
Mappings
(SNPs,
refseqs,…)
Mappings
(SNPs, refseqs,…)
Reference genome:
Sequencing
HiSeq 2000 2x100
paired end:
Mapping (BWA):
SNP identification:
SNP filtering:
Sequencing
76 animals
Sequencing
98 animals
Loss of function SNPs
Allele frequency in
NZ sheep breeds
32M 18M
(MAF >=0.1)
6M
(SNP proximity
filter)
3M
(design score >=0.9
Infinium II)
564 k
Equal spacing
5M
(MAF>=0.02)
Functional
GBS
Literature
50k Chip
55 K
SNPs located on the 50K SNP chip
0.3k*2
19k
685k
695k
Total SNPs
Sheep HD SNP chip
SNP category: ~SNP #:
+
(SNP proximity and design score filters )
(SNP proximity and design score filters )
48k
(SNP proximity and design score filters )
Total probes
Sheep HD SNP chip
Repeatability of the call rate was
99.9978
Mendelian inheritance
ave 651 failures out of 600,500 SNPs
called for each of the 128 trio
comparisons available in the QC set
i.e. 99.89% of SNPs passed this
criteria.
Failed SNPs
lower call rates (would have been
excluded anyway).
most common reason: homozygote calls
when in fact it was a heterozygote.
usually caused by a nearby variant on
one allele interfering with oligo binding.
Oligo probes for
685, 734 SNPs
606,006 SNPs passed
manufacturing QC (87%)
NZ populations:
603,350 SNPs
mean call rate of 99.75% across SNPs
and animals.
NZ populations:
589,630 SNPs
threshold >98% of animals have
to have a genotype
MAF in New Zealand populations.
536,373 assayable SNPs (ie MAF>5%) for inclusion into genomic selection analyses
similar or higher than those typically reported in cattle for GS
considerably in excess of the predicted requirement of 350,000 SNPs for across
breed prediction based on linkage disequilibrium
Sheep HD SNP chip
-
50,000
100,000
150,000
200,000
250,000
0.1-0.19 0.2-0.29 0.3-0.39 0.4-0.5
count
MAF
MAF bins of the “equal spacing”
SNPs MAF in NZ populations
Use of HD chip in NZ
FarmIQ
Concentrated on NZ Terminal sires & meat quality traits
No genomic predictions available from elsewhere
Meat quality really only viable in terminal sires
Beef + Lamb New Zealand Genetics
dual-purpose traits affecting on-farm productivity
PGgRc-NZAGRC (Pastoral Greenhouse Gas Research Consortium
and the Agricultural Greenhouse Gas Research Centre)
investigating the genetics governing sheep methane emissions.
Phase 2: Trait Investigation
Identify and measure eating quality and carcass yield traits in
progeny of terminal sires.
Animals processed through Silver Fern Farms plants; utilise the SFF
system of automatically tracking carcasses based on the use of
electronic identification (EID).
Eating quality indicators :meat tenderness, pH, marbling, colour and
colour stability.
Working with 11 New Zealand flocks that produce Texel terminal
sires or composite-type genetics derived from the Suffolk, Texel,
and Poll Dorset breeds; most of them part of the Focus Genetics
network.
FarmIQ
animals with meat yield (carcass weight, GR, Butt
circumference)
and
quality measurements (pH, colour, tenderness, marbling)
Flock YOB Sires Yield Quality NIR EQ loins
collected StdSFF
Progeny test A 2010 186 2369 2301 2369
2011 105 2396 1169 425 2396
2012 77 1181 1168 580 1181
2013 118 1199 1187 1187 1187 1199
2014 105 1200 1200 1200 1200
Total 581* 8345 7025 3392 1187 8345
Progeny test B 2010 55 1319 623 1319
2011 43 1336 655 108 1336
2012 37 1134 557 1010 1134
2013 47 1166 567 567 567 1166
2014 42 922 500 500 40
(20/slaughter) 1000
Total 223* 5877 2902 2185 607 5955
Progeny test C 2013 20 422 293 293 293 422
2014
20 (12 MQ
+8 dual) 550 300 300 20 550
Total 38* 972 593 593 313 972
Monitor farm A 2011 24 592
2012 15 949
2013 10 695
2014 10 20 800
Total 29* 20 3036
Monitor farm B 2013 10 434
2014 6272
Total 16 706
Grand
Total 839* 15,194
10,520
6,170 2,127 19,114
Genetic parameters-Summary
Traits are heritable (0.1-0.5)
Meat quality and carcass traits especially
Variation in growth CV ~14% of mean
There are breed (or flock) differences
Variation in meat yield and quality traits
Typically same or higher up to CV 30% (except pH)
Phase 3: Genotyping, GWAS and mBVs
Genotype resource
~10,000 animals born 2010-2012 (flocks A, B) genotyped HD chip (includes
sires)
~5400 animals born 2012-2013 genotyped LD & HD chip
Phenotype resource
~85 traits available
focused > 1,000 phenotype records
Meat quality
Tenderness (KgF)
Marble score (visual)
pH
Colour stability at 8 weeks (24, 48, 96, 168 h)
Model
fixed effects: flock, year of birth, breed, mob and sex
covariate: carcass weight (only for meat quality)
Derived genomic relationship matrix (20,000 autosomal SNPs) and used to fit a
polygenic model
Residuals used in a test for association with each marker with simple linear
regression
Tenderness (KgF) Marbling-visual Colour a (8 weeks 48 h)
Fat depth Eye muscle depth Carcass leg length
Weaning weight
pH
How big?
Tenderness
mean 6.1 kgSF, SD 2.2kgSF, n=4798 animals
MAF =0.44, 1 copy =0.24, 2 copies 0.48 kgSF,
6.2% genetic variation (r=25%)
Best variant located in middle of gene
Marbling Score
Mean 3.1, SD 0.67, n =6011 animals
MAF =0.36, 1 copy 0.05, 2 copies 0.1
5.5% genetic variation (r=23%)
Genomic Selection
Principal components for the FarmIQ animals
genotyped and analysed in GS.
IBS group
o 1
D2
+3
x4
100, 000 SNPs were randomly selected from the 573,028 SNPs to investigate further for GS.
MAF of SNPs in GS
Accuracies of Genomic selection for lamb meat quality traits.
Loin Trait
Ibs
group
h2
Colour (l8a168)
1
0.03
Colour
(l8a96)
1
0.16
Colour (l8a48)
1
0.13
Colour (l8a24)
1
0.16
Shear Force (
KgF)
1
0.18
Marbling Score
1
0.26
pH
1
0.10
Colour (l8a168)
All
0.03
Colour
(l8a96)
All
0.16
Colour (l8a48)
All
0.13
Colour (l8a24)
All
0.16
Shear Force (
KgF)
All
0.18
Marbling Score
All
0.26
pH
All
0.10
GBLUP method used
Validation n=300
Correlations are the mean from resampled training and validation sets, 'iterations'-
fold.
mBV accuracy = cor(tBV,mBV))= cor(Y, mBV)/sqrt(h2)
h2 is the estimated heritability
Phase 4: Testing in commercial flocks
test results in commercial terminal sire breeding flocks to ensure the
system is accurate, practical and cost-effective.
Assisted by the 5K LD SNP (practical to use as part of an expanded
parentage test)
new LD chip build underway ~15K SNPs
Breeders will use the chip to assess their weaner ram lamb crop in early
2015.
~2500 born 2014 ram lambs (as part of existing DNA parentage)
Used industry mating autumn 2015
This work complements other approaches to improve eating quality,
including enhanced pre and post slaughter management.
Summary
Meat quality & eating traits heritable
Genomic prediction accuracies from DNA up to
50%
Looks like some loci with major effects
Can predict across breeds
Probably only explain moderate proportion of
variation
May have to fix or discard if effects both positive
and negative
Next steps
Taste panel to “ground truth” physical measurements
Create meat quality value index ranking animals
Progeny test
Sires from new breeders (4 breeding groups)
More maternal sires (various)
Better linkage between flocks = better breed
rankings
Beta test in industry continued
Timeframes
Year
1 2 3 4 5 6 7
Sequencing
HD SNP
chip
creation
Genotyping
Βeta Test
Commercial use
First crop
FARMIQ
lambs
slaughtered
10,000 lambs for
prediction
equation
10,000 lambs +
Prediction Eqn
developed
Elite &
Commercial
Sires used
with FARMIQ
Eqn
Meat trait measurement - manual Meat trait measurement - auto
Acknowledgements
John McEwan
Michael Lee
Rudi Brauning
Wendy Bain
Rayna Anderson
Tracey Van Stijn
Ken Dodds
Animal Genomics Team
ResearchGate has not been able to resolve any citations for this publication.
ResearchGate has not been able to resolve any references for this publication.