Transcript Researcher school genetics / Forskarskule genetikk

RESEARCH SCHOOL GENETICS
NORWEGIAN UNIVERSITY OF LIFE SCIENCES
www.umb.no
RESEARCH SCHOOL GENETICS
/
FORSKARSKULE GENETIKK
Tormod Ådnøy, leader
5.9.2007
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Welcome!
Program today
 1415 On the Research school genetics at UMB. Tormod
Ådnøy
 1438 ’Lille lørdag’ – Åsmund Bjørnstad
 1440 Group work – Who we are, what we know, and
what we want. Groups of 3-5 participants.
 1455 Brief summing up by the youngest in every group.
 1500 Pizza, beer, .. (free) (Husdyrkantina)
Discussions over tables, and all together. (Genetic small talk
permitted)
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
What we are
 An application for funding
(web - in Norwegian) /
100.000 nok
But people and ideas are more important than money:
 Email list: 42 PhD, 45 Advisors and researchers
 Web page http://www.umb.no/22912
 Preliminary board (3 PhD, 2 advisors)
– Silje Brenna Hansen (PhD Cigene)
– Marianne Haraldsen (PhD IHA – Forskargruppe genetikk og avl)
– Simen Rød Sandve (PhD IPM – Genetikk og plantebiologi)
– Morten Lillemo (postdoc IPM)
– Tormod Ådnøy, leader research school (assoc.prof. IHA)
 A secretary: Anne Golten, IHA
 This gathering today, and first Wednesday every month
Focus on PhD students
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
What we will be remains to be seen …
 Send an email for new participants to join the school
– So far membership is not exclusive
– Meetings are open
 Peer review groups?
 Reader groups?
 Nordic collaboration?
 Research grant applications?
 Include MSc students?
 ECTS for some activities?
 Presentation of own work for others in the Research school
– May help self-image
– Will give useful training
 Future courses in the Research school?
 Summer courses?
 ...
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
What I will talk about now:
 ‘SCHOOL’
•
•
•
•
•
 ‘GENETICS’
Institution
Knowledge
Feeling ok
Clowns
Paradigms
– Genes
• DNA, mRNA, ..
• SNP
• Genotypes,
haplotypes
• Regulatory nets
– BREEDING
• Regression
• Additive relationship
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
GENETICS
 Gene
– Gene maps, DNA, mRNA, … , amino acids and proteins, …
– SNPs identifying genes in single individuals
– We know a lot more now than some years ago
Molecular lab people have a lot of information – and will have a lot
more!
How can it be used?
Can it be used to find the best future individual for a trait we want
to improve?
What combination of genes is best?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
How is the gene expressed in a trait?
We don’t see a gene’s value, we see an individual’s complete
genome’s value!
 Genotype value for the trait
How do we express a gene’s value
– or
How do we know which genes to combine to have a better
individual in the future?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
If two individuals were the same except two alleles in a locus
We could say that the difference in the two individuals’ genotype
values was the difference of the two allele effects
But the alleles may interact with other genes, or the environment
And normally we have a lot more differences between two
individuals than just two different alleles
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Numbers of genotypes
 ..very many
 How do we know which individuals/genotypes to select
for future breeding?
– What is you answer?
 May we predict what value a not yet existing genotype
(of infinitely many) will get for a trait?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
.. to have better individuals in the future ..
 Select for single gene effects
 Select for haplotype effects
 Select for combination of gene effects (dominance,
epistasis, heterosis)?
 Select for best genotypes today = breeding
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Casein genes in Norwegian goats
 DNA from 436 bucks in national breeding scheme
– Analyzed 39 snps (single nucleotide polymorphisms) in 4
casein genes (on same chromosome)
 Haplotypes deduced from snp genotypes and relationship
 Milk (kg), and protein-, fat-, lactose-% from daughters in
Goat dairy control
» Hayes,Ben; Hagesæther,Nina; Ådnøy,Tormod; Pellerud,Grunde;
Berg,Paul R.; Lien,Sigbjørn (2006): Haplotype structure of casein
genes in Norwegian goats and effects on production traits.
Genetics 174, 455-464.
www.umb.no
1 2 3 4 5 6 7 8 9
1996043
1996446
1996846
1997305
1997769
2003668
2003670
2003671
2003673
2003674
2003841
2003842
2003843
2001181
2001182
2001185
2001186
2001187
2001213
2001232
1997782
1998307
1998429
1998450
1998456
1998590
1998607
1998735
1998745
1999093
A C
A C
A CT
A CT
A C
A C
A C
A C
A C
A CT
A ?
A C
A C
AG C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
A C
? ? AG TC
GA AG AG TC
GA ? AG TC
GA AG AG TC
G G G C
G G G C
G G G C
GA ? AG TC
G ? G C
GA ? AG TC
GA ? AG TC
GA ? AG TC
G ? G C
? ? AG TC
? AG AG TC
G G G C
G G G C
G G G C
G G G C
GA ? AG TC
G ? G C
G G G C
G G G C
G G G C
G G G C
G G G C
G G G C
G G G C
G G G C
G ? G C
AG CT
AG CT
AG ?
AG CT
G T
G T
G ?
AG ?
G T
AG CT
AG ?
AG CT
G ?
AG CT
AG CT
G T
? ?
G T
G T
AG CT
G T
G T
G T
G T
G ?
? T
G T
G ?
G T
G T
10
? DEL.C
? DEL.C
? C
? C
C ?
C C
C C
CG DEL.C
C C
? C
? C
? DEL.C
C C
? ?
? DEL.C
C C
C C
C C
C C
CG DEL.C
C C
C C
C C
C C
C C
? C
C C
C C
C C
C C
11 12 13
CSN1S1_E12-0
CSN1S1ex17_16860
CSN2exon7_11801
CSN2exon7_11770
CSN2prom_2071
CSN2prom_1653
CSN2prom_1009
CSN2prom_862
CSN2prom_760
CSN1S2ex16_273
CSN1S2ex16_682
CSN1S2ex16_987
CSN3Prom_677
CSN3Prom_833
CSN3Prom_852
CSN3Prom_942
CSN3Prom_991
CSN3Prom_1074
CSN3Prom_1140
CSN3Prom_1191
CSN3Prom_1338
CSN3Prom_1499
CSN3Prom_1550
CSN3Prom_1935
CSN3Prom_2136
CSN1S1ex9_9889
CSN1S1in9_9918
CSN1S1ex10_10673
CSN1S1_E12-F3
Buck
CSN1S1prom_264
CSN1S1prom_866
CSN1S1prom_888
CSN1S1prom_1105
CSN1S1prom_1169
CSN1S1prom_1379
CSN1S1prom_1470
CSN1S1ex4_6075
CSN1S1ex4_6091
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Casein snp genotypes – excerpt of the 436 bucks
14
AG CG A A
AG CG A DEL.A
AG CG GA DEL.A
AG CG GA DEL.A
A C A DEL
A C A DEL
A C A DEL
AG CG A DEL.A
A C A DEL
AG CG GA DEL.A
AG CG GA DEL.A
AG CG A DEL.A
A C A DEL
AG CG GA DEL.A
AG CG A A
A C A DEL
A C A DEL
A C A DEL
A C A DEL
AG CG A DEL.A
A C A DEL
A C A DEL
A ? A DEL
A C A DEL
A C A DEL
A C A DEL
A C A DEL
A C A DEL
A C A DEL
A C A DEL.A
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38
C
C
TC
TC
C
C
C
C
C
TC
TC
C
C
TC
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
T
T
CT
CT
T
T
T
T
T
CT
?
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
T
?
C
C
C
C
C
C
C
C
C
C
C
C
?
C
?
C
C
C
C
C
C
C
C
?
?
C
C
C
C
A G
A G
GA G
GA G
A G
A ?
A ?
A ?
A ?
GA ?
GA ?
A ?
A ?
? G
A G
A G
A G
A G
? G
A ?
A G
A G
A G
A G
A G
A ?
A ?
A G
A G
A G
www.umb.no
AG TA
AG ?
A T
A T
A T
A T
A T
AG TA
A T
A T
A T
AG TA
AG TA
A T
AG TA
A T
A T
A T
A T
AG TA
A T
A T
A T
A T
A T
? TA
A T
A T
A T
A T
TC
TC
C
C
C
C
C
TC
C
C
C
TC
TC
?
TC
C
C
C
C
TC
C
C
C
C
C
TC
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
C
?
?
C
C
CT
CT
C
CT
T
C
C
CT
C
?
C
CT
CT
C
CT
C
CT
CT
CT
CT
CT
C
C
CT
?
CT
?
AT
A
A
AT
AT
A
AT
T
A
A
AT
A
AT
A
AT
AT
A
AT
A
AT
AT
AT
AT
AT
A
A
?
AT
AT
G G G A
G G G ?
GA GA G AT
A A G T
GA A G T
GA ? G AT
A ? G T
G G G A
G ? G A
A ? G T
A G G T
G ? G A
G ? G A
? GA ? AT
G GA G AT
GA GA G AT
GA GA G AT
G G G A
G G G A
GA ? G AT
? G G A
GA GA GA AT
GA GA G AT
GA GA G AT
G G G A
GA G G AT
GA ? G AT
GA GA G AT
GA GA G AT
GA GA G AT
C
?
TC
T
TC
TC
T
C
C
T
T
C
C
TC
C
TC
TC
C
C
TC
C
TC
TC
TC
C
TC
TC
TC
TC
TC
?
T
TA
A
TA
TA
A
T
T
A
A
T
T
TA
T
TA
TA
T
T
TA
T
TA
TA
TA
T
TA
TA
?
TA
TA
G T
G T
? TG
C ?
C TG
GC TG
C TG
G T
G T
C ?
C TG
G T
G T
GC TG
GC T
GC TG
? TG
G T
G T
? TG
G T
? TG
? TG
GC TG
G T
? TG
GC TG
? TG
GC TG
? TG
T
T
TG
G
G
TG
G
T
T
G
G
T
T
TG
T
TG
TG
T
T
TG
T
TG
TG
TG
T
TG
TG
TG
TG
?
A TC
A TC
A T
A T
A T
A TC
A T
A TC
A C
A T
A T
A TC
G T
A TC
A T
A TC
A TC
AG T
A T
A T
A TC
A T
A TC
A TC
AG TC
AG T
A T
A TC
A TC
A TC
GT
?
G
G
G
GT
G
GT
T
G
G
GT
G
GT
G
GT
GT
G
G
G
?
G
GT
GT
GT
G
G
GT
GT
GT
G
G
G
G
G
G
G
G
G
G
G
G
A
?
G
G
G
GA
G
G
G
G
G
G
GA
GA
G
G
G
G
0,08
0,06
Effect on DYD
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Results haplotypes – effects on fat-, protein-% and
milk, not significant for lactose%
0,04
9 11
0,02
19 17
6 1
12
10
3
15
Fat %
0,00
-0,02
Prot %
18 21
-0,04
14
8
4
13
2
-0,06
-0,08
Haplotype
www.umb.no
20 7
5
Milk
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Results of single snp – not significant
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Finding gene effects / Number of genotypes
 We deduced 21 haplotypes on bucks, and found additive
effects on daughters’ production.
 All possible combinations of 39 snps is 339>1018, but
number of individuals observed was 436. All genotypes
may not be modeled, only the ones observed.
 Modeling additive effects of all 39 snp-s simultaneously
led to collinearity problems, but we could analyze for one
snp at a time.
 Even to find all haplotype combinations represented in a
sample will be difficult: 21+21*20/2=231 potential
genotypes. (Some haplotypes are rare.) How important is
haplotype dominance?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
BREEDING
 To generate the best future individuals – We want to
change the population mean
– Info used:
• Phenotypic observations
• Additive relationship
..best genotypes/ population (for a future environment)
– Given
• Existing populations,
• Existing knowledge about the populations,
• Existing techniques for breeding (AI, blup, ..)
 Focus is on population, less on individuals
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
How do we know which individuals/genotypes to
select for future breeding?
– Select best phenotypes – Mass selection
– Select individuals with best offspring –
– Other methods
Breeders use genes as an alibi – they don’t need them!
Statistics: linear regression of offspring phenotypes on parents’
phenotypes
Additive inheritance of genes is a motive for relationship matrix
– Include info on genes
» Meuwissen, T. H. E., Hayes, B. J., & Goddard, M. E. (2001).
Prediction of total genetic value using genome-wide dense marker
maps. Genetics 157, 1819-1829.
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
The infinitesimal model
 Many genes
 Small effects
 Independently distributed
 No change in gene frequency
 Equilibrium of gene frequencies
All assumptions are violated in breeding programs, normally
‘Shaky foundation of Fisherian genetics’ – SWO
 Why does it work so well?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Additivity
A crucial question:
 To what extent are gene effects additive?
 How does deviations from additivity affect the Parent-
Offspring relationship:
Cov (P,O) =0.5*Additive variance ?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Genotype values
Two additive loci (aH=2, aK=3)
10
9
8
7
6
5
4
3
2
HH
1
Hh
0
KK
hh
Kk
kk
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Genetic variance – pure additive model
 For two loci with two alleles each (H h K k), and only additive gene effects (in figure:
aH=2 aK=3, while ah=ak=0):
 Let Hi=1 when H-allele is present, Hi=0 when h-allele
 Then the genotype value is
y(i,j)= [H(i)+H(j)]*aH + [K(i)+K(j)]*aK
(0, 2, 4, …, 10 in figure)
 The mean genotype value is
EY= sum p(i,j) * y(i,j) = 2*pH*aH + 2*pK*aK
 The variance of the genotype values, with random mating and same
disequilibrium in parents’ gametes (’D’= dHK=pHK-pH*pK)
VY= E(Y2)+(EY)2= 2*pH*ph*aH2+2*pK*pk*aK2+4dHK*aH*aK
= VY0 + VYd
Avery, P. J. & Hill, W. G. (1978). The effect of linkage disequilibrium on the genetic variance of a quantitative trait.
Adv. Appl. Prob. 4-6. /
Ådnøy, T. (1981). Selection in few-locus models / Seleksjon i få-lokus modellar. PhD-dissertation at Dept Mathem
Statist, Agric Univ Norway. 1-218.
 Even in the additive model, disequilibrium over loci will change the
variance.
Linkage, selection, .. may lead to disequilibrium.
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
 If selection is for an additive trait we should expect that
the best allele is fixed in every locus
– Should be no genetic variation left
 This does not happen normally
– There is genetic variation left for most traits even after
much selection
– Why?
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Bridging the gap: genes – phenotypes (Cigene in eVita)
Arne Gjuvsland (Cigene) PhD dissertation October 2:
 linking regulatory gene networks to additivity and
dominance
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
SCHOOL
 To learn
– Something is not known by students
– It is normal that students don’t know – that’s why they attend a
school
The most important is the process in the students’ heads
 Transfer of knowledge – from lectures, books, ..
– It helps to know what you already know
 Generation of knowledge
– Important science may generate new ’schools’ (paradigms)
» ’The shaky foundations of Fisherian genetics’ SWO
 Creativeness is good in a research school / new ideas
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
To feel OK
 Emotions are important for learning
» We learn more when we fell ok
 We (most of us) need to know if what we are doing is
good/relevant/useful/
– We may not always rely on our self-evaluation
– We need external evaluations
» Norwegians are good at belittling themselves
– We need to compare to what others do
 What do you need to trust that you are doing ok?
• If I tell you you’re clever – do you believe me?
 Others’ input may correct our learning – make us better
students
 Don’t be afraid to tell what you don’t know!
– Helps other feel helpful / builds their self-image
 Clowns help us relax
– May help us see ourselves in a new light
www.umb.no
 Genetics
 Breeding
 School
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
I have talked about
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Litle laurdag / Lille lørdag
Wednesday=little Saturday
 Now professor Åsmund Bjørnstad will tell a story?
Fanfare!!
In comes the clown??
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
www.umb.no
– Divide in groups
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Groups
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
GROUP ASSIGNMENT
 Who we are
– Present yourself to the group
• Name, birth, occupation, …
 What we know
– What techniques do you use? What courses are you taking?
• Variance components, Linear models, Molecular lab, mRNA,
micromatrices, HFA401, …
– How does your discipline find the ’best’ individuals for the future?
 What we want
– How can a research school be useful?
– What can we contribute yourself and what can we get/buy from others?
– Present two topics where you think our school may be helpful to the
whole group at 1455. (By youngest in group.)
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
Food and drink
 Pizza
– 8 assorted kinds
– 1 vegetarian
– 1 ’muslim’
 Salad with vinaigrette
 One bottle of drink
– Apple drink
– Clausthaler Beer without alcohol
– Green Tuborg
I need two voluntaries to help with the dishes afterwards
www.umb.no
RESEARCH SCHOOL GENETICS NORWEGIAN UNIVERSITY OF LIFE SCIENCES
www.umb.no