zehn jahre genomische selektion in der schweiz...bulldog, doppellender hornlos, farbe. | 30...
Post on 24-Aug-2020
0 Views
Preview:
TRANSCRIPT
www.qualitasag.ch | 1
Zehn Jahre genomische Selektionin der Schweiz
Plattform der Rassenclubs
Mutterkuh Schweiz
15. Dezember 2017
juerg.moll@qualitasag.ch
www.qualitasag.ch | 2
Ein zentrales Konzept in der Tierzucht...
Phänotyp Genotyp Umwelt
• Leistungen• Exterieur• Fitness• Verhalten
• Erbgut • Haltung• Fütterung
Grundlagen
www.qualitasag.ch | 3
Verbesserung der genetischen Veranlagung der Tiere für bestimmte Leistungs- und Fitnessmerkmale
Kenntnis des Erbgutes von Zuchttieren möglichst:• früh• genau• kostengünstig
?
?
Ziel der Tierzucht
www.qualitasag.ch | 4
Der „klassische“ Weg:
• Leistungsprüfung (Absetzgewicht, Schlachtdaten usw.)
• Herdebuchführung
• aufbauend darauf Zuchtwertschätzung
Der neue Weg:
• Leistungsprüfung (Absetzgewicht, Schlachtdaten usw.)
• Herdebuchführung
• aufbauend darauf Zuchtwertschätzung
• Informationen aus dem Erbgut
= genomische Selektion
Genetische VeranlagungWie beschaffen wir uns Kenntnis über das Erbgut?
www.qualitasag.ch | 5
• Direkt nur sehr schwer möglich– Quantitative Merkmale (viele Gene wirken auf ein
Merkmal)– wenig bekannt über Lage und Wirkung der Gene
• Genomische Selektion arbeitet mit SNP-Markern• SNP: Single Nucleotide Polymorphism
– punktuelle Veränderungen im Genom (1 bp lang)– kommen in sehr hoher Anzahl vor– nur je zwei Varianten
• Marker zeigen uns Genwirkungen aufgrund der physischen Nähe zu den Genen– Kopplungsungleichgewicht
Gene entdecken – WIE?
www.qualitasag.ch | 6
SNP1
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
AAABBB AAAB BBBB AA ABBB
SNP2
SNP3
SNP4
SNP5
SNP6
SNP7
SNP8
SNP10
SNP9
.........
Effektschätzung?
www.qualitasag.ch | 7
1) Vergleich von Zuchtwerten mit Genotypen
– Datengrundlage für die SNP-Effektschätzung bilden die SNP-Typisierungen und die konventionellen Zuchtwertevon Stieren* mit einem Nachzuchtprüfungsresultat
– Trainingsdatensatz (Referenzdatensatz, Kalibrierungsdatensatz, Lernstichprobe)
2) Schätzen der genomischen Zuchtwerte der anderen Tiere mit der Formel aus 1)
* teilweise werden auch Kühe im Trainingsdatensatz verwendet; diese sollten zufällig aus der Population ausgewählt werden
Effektschätzung - Prinzip
www.qualitasag.ch | 8
kg M
ilch
BBABAA
Beispiel mit einem SNP:
SNP-Effekt
Der Effekt eines zusätzlichen „B“-Allels beträgt +10kg
Effektschätzung
www.qualitasag.ch | 9
Die Berechnung der genomischen Zuchtwerte ist sehr einfach, denn man muss nur SNP-Effekte aufsummieren
SNP1
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
CAGTATCGTAATAATAATAATAAGTCCTGGGCCTAAACCTGCAAGGTTTCGATATTGGCCCCCGG GTCATAGCATTATTATTATTATTCAGGACCCGGATTTGGACGTTCCAAAGCTATAACCGGGGGCC
Kalb 1
AAABBB AAAB BBBB AA ABBB
SNP2
SNP3
SNP4
SNP5
SNP6
SNP7
SNP8
SNP10
SNP9
+1.3 -0.1 +0.5 +0 -0.4 +3.8 +38.3 +4.1 -5.7 +2.7
ZW = 2*(+1.3)+1*(-0.1)+1*(+0.5)+0*(0)+0*(-0.4)+ 2*(+3.8)+2*(+38.3)+0*(+4.1)+2*(-5.7)+1*(+2.7)
ZW = 81.2 = direkter genomischer Zuchtwert (DGZW)
Berechnung genomischer ZW
www.qualitasag.ch | 10
20152010200520001995
Trainingsstiere (nachzuchtgeprüfte Stiere)Selektions-Kandidaten
EffektschätzungSumme derSNP-Effekte
(DGZW)
Effektschätzung
www.qualitasag.ch | 11
2005
Trainings-Stiere
EffektschätzungSumme der
SNP-Effekte (DGZW)
Validierungs-Stiere
Validierung:
Korrelation zwischen DGZW
und trad. ZW
2015201020001995
Sicherheit genomischer ZW: Validierung
www.qualitasag.ch | 12
Merkmal BV RH/HO
Produktion (Ekg) 49 % 56 %
Zellzahl 43 % 44 %
LBE 43 %
(25 - 56) 43 % (21 - 61)
Fruchtbarkeit 38 % 35 %
Nutzungsdauer 27 % 42 %
Sicherheit (B%) der DGZW
www.qualitasag.ch | 13
Zuchtwert (ZW): „konventionell“ geschätzt, ohne Einbezug von Markerinformation
Direkter genomischer Zuchtwert (DGZW): Zuchtwert geschätzt allein aufgrund von Markerinformationen
Genomisch optimierter Zuchtwert (GOZW): Zuchtwert geschätzt auf Grund von traditionellen Daten und Markerinformationen(Kombination von ZW und DGZW).
Begriffe
ZW-Typ traditionell Deklaration GOZW Tiergruppe
Abstammungs-ZW+ DGZW
GA Jungtiere
CH-Zuchtwert G Kühe, Stiere
Interbull-Zuchtwert GI Stiere
www.qualitasag.ch | 14
−20 0 20 40 60 80
−20
020
40
60
80
konv. ZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.54
−20 0 20 40 60 80
−20
020
40
60
80
DGZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.7
−20 0 20 40 60 80
−20
020
40
60
80
GOZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.72
Genauigkeit genomischer Zuchtwerte
Vergleich ZW Eiweiss kg 2012 - 2016
−20 0 20 40 60 80
−20
020
40
60
80
konv. ZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.54
−20 0 20 40 60 80
−20
020
40
60
80
DGZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.7
−20 0 20 40 60 80
−20
020
40
60
80
GOZW Eiweiss kg 2012
konv.
ZW
Eiw
eis
s k
g 2
016
Korrelation: 0.72
Abstammungs-ZW GOZW
kon
v. Z
W m
it N
ach
kom
me
n r = 0.54 r = 0.72
www.qualitasag.ch | 15
Anzahl Trainingsstiere Braunvieh
• Je grösser der Trainingsdatensatz desto genauer die genomischen Zuchtwerte
• Kühe im Trainingsdatensatz?• Heute ziemlich vollständiger Trainingsdatensatz bei
Braunvieh (Projekt Intergenomics): ca. 6‘500 Stiere Ekg
Aug 2011 Dez 2011
n Training corr(DGZW,CHZW)n = 250
n Training corr(DGZW,CHZW)n = 360
Veränderung (%)
Mkg 2471 0.60 3300 0.65 + 4.7
Fkg 2162 0.60 3202 0.70 + 10.6
Ekg 1710 0.53 3254 0.65 + 11.2
ZZ 2188 0.70 2896 0.70 ± 0.0
www.qualitasag.ch | 16
Anzahl Trainingsstiere Holsteinweltweit 2 grosse „Blöcke“
Warum hatte Schweiz lange keinen Zugang?
• kleine Populationen sind nicht sehr interessant
• Offener Zugang zu Typisierung von Stieren war hinderlich
CDDR (erweitert) EuroGenomics
USA Frankreich
Kanada Deutschland
Italien Die Niederlande
Grossbritannien Nordische Staaten
Schweiz (seit 2016) Spanien
Polen
www.qualitasag.ch | 17
• CDDR (Cooperative Dairy DNA Repository)
– Genotypen-Pool von KBO aus USA & CAN
• Erster Austausch im Februar 2016
– SNPs von 125‘000 HOL-Stieren erhalten
– 260‘000 Tiere im Pedigree erfasst
– 14 verschiedene SNP-Chips
– SNPs von 6000 HO/RH-Stieren geliefert
• Monatliche Updates
– Neu typisierte Stiere
Genotypenaustausch Holstein
www.qualitasag.ch | 18
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●●
●
●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
● ●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●●
●●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●●●●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
● ●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●●
●
●●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●●
●●
●
●
●●
●
●
●
●
●
●
●
●●
●
●●
●●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●●
●
●
●●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●●●
●●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●● ●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●●●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●●
●
●
●●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●● ●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●●
●●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
●●
●
●
●
●
●●
●
●
●
●●
●
●
●
●
●
●
●
●●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●
●●
●
●
−40 −20 0 20 40 60 80
−40
−2
00
20
40
60
80
DGZW ORG vs Wonderment
DGZWs ORG
DG
ZW
s W
on
derm
en
t
Mean24.82
Stdv
15.41
Mean
24.69
Stdv
15.4
Korrelation: 1Rangkorrelation: 1
DGZW (Effektschätzung OHNE Stier XY)
DG
ZW
(E
ffe
ktsc
hät
zun
g M
ITS
tie
r X
Y)
Stier XY selbstNachkommen
Enkel
Zusammensetzung Trainingsstiere (Bsp. BV Training mit / ohne Stier XY)
Enger Bezug zwischen Trainingsdatensatz und Selektionskandidaten!
Training - Selektion
www.qualitasag.ch | 19
Fazit:Genotypen sind Spiegel-bild der Population
Simmental hat kaum genetische Verknüpfung zum Trainingsdatensatz:Validierung für viele Merkmale ungenügend
Struktur Trainingsstiere (HO/SI)
www.qualitasag.ch | 20
• Original Braunvieh funktioniert mit gemischtem Trainingsdatensatz (mit BV/BS)
– Stiere max. 95% BS
– Kühe mit OB-Anteil > 50% erhöhen Genauigkeit
• Simmental: nur Milch + Zellzahl (ITB-ZW), übrige Merkmale/Fleisch in Entwicklung
• Erhöhung der Anzahl SNP‘s (50‘000 800‘000) verbessert die Genauigkeit nicht
Genomische Selektion für kleine Populationen
www.qualitasag.ch | 21
X X XMachbarkeit Limousin
Aktuelle Anzahl Limousin-Stiere mit Sicherheit Zuchtwert > 50% und DNA
• Qualität der DNA muss in Ordnung sein• Trainingsdatensatz mit 1’500 Stiere = Investition CHF 150’000.-
www.qualitasag.ch | 22
Machbarkeit Angus
X X XXX
Aktuelle Anzahl Angus-Stiere mit Sicherheit Zuchtwert > 50%
www.qualitasag.ch | 23
Internationale Zusammenarbeit
• Nur mit Win – Win - Modell• Internationale Kooperation wichtig
(Interbeef, Genotypenaustausch)• Länder-/Populationsspezifische
Anpassungen/Entwicklungen notwendig• Herausforderung: Neue Methoden für
kleine Rassen nutzbar machen• Kommerzialisierung von Phänotypen,
Genotypen und Methoden• je mehr „nicht-öffentliche“ Investitionen in
Zuchtprogramme erfolgen, desto grösser wird das Interesse der Investoren kommerzielle Zuchtunternehmen
www.qualitasag.ch | 24
Höherer Zuchtfortschritt ...
• dank genaueren Zuchtwerten (v.a. Jungtiere) und kürzerem Generationenintervall
• weil „schwierige“ Merkmale besser züchterisch bearbeitet werden können (z.B. Fruchtbarkeit)
• weil neue Merkmale züchterisch bearbeitet werden können (z.B. Gesundheitsdaten)
Zucht mit genomischer Selektion
www.qualitasag.ch | 26
Gesamtzuchtwert der Stiere aus Genotypenaustausch Holstein
- Gesamtzuchtwert ISET- Inzuchtgrad F
www.qualitasag.ch | 27
Inzucht
• Inzuchtdepression = verminderte Leistungsfähigkeit und Fitness
• Auftreten von Erbfehlern
• Verminderung der genetischen Varianz = Gefährdung zukünftiger Zuchterfolge
www.qualitasag.ch | 28
Optimum genetic contribution
• optimiert den Zuchtfortschritt bei definiertem Inzuchtanstieg
• Sucht Tiere mit hohen Zuchtwerten und tiefem durchschnittlichen Verwandtschaftsgrad zur Population
www.qualitasag.ch | 29
Zusatznutzen
• Abstammungskontrolle Mutterkuhrassen AN, BV, DR, LM und SM
• Erbfehler/ZusatztestsBulldog, DoppellenderHornlos, Farbe
www.qualitasag.ch | 30
Erbfehler dank SNP‘s (Auszug)
BH1 Embryonaler Fruchttod BS
BH2 Lebensschwache Kälber BS,FV
CD Durchfall HO
FH1 Zwergwuchs FV
FH2 Minderwuchs (Leberproblem) FV, OB
FH4 Embryonaler Fruchttod FV
FH5 Herzschwäche, Tod innert 48h pp FV
HH1 Embryonaler Fruchttod HO
HH2 Embryonaler Fruchttod HO
HH3 Embryonaler Fruchttod HO
HH4 Embryonaler Fruchttod HO
HH5 Embryonaler Fruchttod HO
JH1 Embryonaler Fruchttod JE
MH1 Embryonaler Fruchttod MO
MH2 Embryonaler Fruchttod MO
www.qualitasag.ch | 31
Erbfehlerstrategie
• Viele neue Erbfehler entdeckt
• Neue Strategien erforderlich
– Ausschluss von Trägertieren nicht mehr für alle Erbfehler sinnvoll
– Vermeidung von Träger x Träger Paarungen (Machbarkeit im Natursprung?)
– Genetic Load Index (Frequenz, Schaden, Vererbung)
www.qualitasag.ch | 32
ChipName
GGPLD v2 (9K)
GGPLD v3 (26K)
GGPLDv4 (30K)
50Kv1 (54K)
50Kv2 (54K)
GGPHD (80K)
GGPHD (150K)
HD (850K)
GGPLD v2 (9K)
8.762 92.9% 92.9% 92.6% 92.3% 92.1% 93.6% 95.8%
GGPLD v3 (26K)
26.151 99.8% 32.1% 31.9% 36.3% 95.0% 98.8%
GGPLD v4 (30K)
30.125 34.1% 35.9% 95.1% 97.2%
50K v1 (54K)
54.001 96.6% 52.4% 77.2% 90.2%
50K v2 (54K)
54.609 51.7% 75.8% 90.5%
GGPHD (80K)
76.999 95.9% 96.8%
GGPHD(150K)
139.481 96.3%
HD(850K)
777.962
SNP Chip Modelle – Überlappung
www.qualitasag.ch | 33
IMPUTING = Auffüllen der fehlenden SNPs bei den LD TierenInformationsquellen:
– Pedigree
– Populationsinformation (Linkage Disequilibrium)
Imputing: Daten
www.qualitasag.ch | 34
durchschn.
% korrekt
durchschn.
% inkorrekt
durchschn.
Korrelation
imputiert-wahr
Beide Eltern 98,6 1,4 0,98
Vater + MGV 98,1 1,9 0,97
Vater 97,6 2,4 0,96
Andere 97,3 2,7 0,95
Imputing LD auf 50k
Daten: 3‘738 mit 50k typisierte BV-Tiere
Annahme: die 723 jüngsten Tiere sind mit LD Chip typisiert
Frage: wie gut werden fehlende SNPs geschätzt?
www.qualitasag.ch | 38
Typisierungen nach Auftraggeber
0
1000
2000
3000
4000
5000
BVCH SHZV shb weitere
2011 2012
2013 2014
2015 2016
www.qualitasag.ch | 39
Anzahl Typisierungen 2016
5679
1028 66
GGP LD
GGP HD
HD
www.qualitasag.ch | 40
Chip Laborpreise GeneSeek
• GGPLDv4 (30K) $ 38.00• 50Kv2 (54K) $ 85.00• GGPHD (150K) $ 85.00• HD (770K) $ 165.00
Preise für SNP-Typisierung inkl. DNA-Aufbereitung
• Probenhandling Qualitas CHF 5.- bis 17.-
www.qualitasag.ch | 41
Durch Wettbewerb / Konsumenten gefordert!
• Gesundheit, Immunität
• Futterverwertung, Effizienz, Treibhausgasemissionen
• Tierwohl, Tierverhalten
• Fleischqualität
Genomische Selektion ermöglicht schnellere Verfügbarkeit!
Genomische Selektion für neue Merkmale
www.qualitasag.ch | 42
• Neue Konzepte für die Leistungsprüfung
• Prüfbetriebe Vertragsbetriebe mit möglichst zuverlässiger Erfassung von züchterisch interessanten Merkmalen
• Genotypisierung von weiblichen Tieren
– Mehrere 1000 Kühe mit guten Phänotypen
• Hohe Investitionen für Züchter
– Internationale Kooperation
Genomische Selektion für neue Merkmale
www.qualitasag.ch | 43
Hohe Investitionen & zunehmende Kommerzialisierung
• Aufbau Trainingsdatensatz / Problem Populationsgrösse / Eigentum & Nutzung SNP-Daten
• Kombination genomische Selektion – Fortpflanzungs-technologien (ET-Stationen & Samensexing)
• Neue Technologien: Genom-Editierung
• Samen von Top-Stieren nur gesext verfügbar
• Weibliche Toptiere im Eigentum kommerzieller Firmen
• kommerzielle genomische Indexe (CLARIFIDE Plus; EvaLim)
Tendenzen Rindviehzucht
www.qualitasag.ch | 44
• Genomische Selektion funktioniert und ist fester Bestandteil der Schweizer Milchviehzucht-programme
• Entwicklung der genomischen Selektion geht weiter, es bleibt noch viel zu tun
• Neue Merkmale (Fitness, Gesundheit, Effizienz) gewinnen an Bedeutung
• Zunehmende Kommerzialisierung der Milchviehzucht
• Fleischrinderzucht kann von den Erfahrungen der Milchviehzucht profitieren
Schlussfolgerungen
top related