Evolutionary genetics of hybrid maize

Post on 22-Jan-2018

276 Views

Category:

Science

4 Downloads

Preview:

Click to see full reader

Transcript

Evolutionary Genetics of Hybrid Maize

Jeffrey Ross-Ibarra www.rilab.org @jrossibarra

Gra

in Y

ield

Year

Gra

in Y

ield

Year

How has breeding affected diversity across the maize genome?

Gra

in Y

ield

Year

How has breeding affected diversity across the maize genome?

How has the genome responded to selection for increasing hybrid yield?

Gra

in Y

ield

Year

How has breeding affected diversity across the maize genome?

How has the genome responded to selection for increasing hybrid yield?

What is the genetic basis of hybrid vigor?

van Heerwaarden et al. 2012 PNAS

99

94

70

137

land races

<1950 inbreds

1960-1970 inbreds

ex-PVP

0

1

2

3

(Oh43, W22, B14)

(B73, 207, Mo17)

10,000 ft view of US Corn Belt

!

IodentNon-Stiff Stalk (NSS)

Stiff Stalk

!

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3an

cest

ry p

ropo

rtio

n

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

IodentNSSSS

AC

TGTG

AC

TCCA

TC

G

A

Inbred 1

AC

TGTG

AC

TCCA

TC

G

A

GTG

AC

TC

TC

G

TC

G

AC

G

AC

TGTG

AC

TCCA

TG

T

G

Inbred 1 2 3

AC

TGTG

AC

TCCA

TC

G

A

GTG

AC

TC

TC

G

TC

G

AC

G

AC

TGTG

AC

TCCA

TG

T

G

Inbred 1 2 3

AC

TGTG

AC

TCCA

TC

G

A

GTG

AC

TC

TC

G

TC

G

AC

G

AC

TGTG

AC

TCCA

TG

T

G

Inbred 1 2 3

AC

TGTG

AC

TCCA

TC

G

A

AC

TGTG

AC

TCCA

TG

T

G

Inbred 1 3

AC

TGTG

AC

TCCA

TC

G

A

AC

TGTG

AC

TCCA

TG

T

G

Inbred 1 33

2

1

0

time

cate

gory

SSNSS

Iodent

SS

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

C

I31A

H

Y

B8

B10

B1

4

B96

B3

7

I205

O

H07

C

I3A

C

I187

-2

CI5

40

I224

W

D45

6

Oh3

167B

Tr

9-1-

1-6

FE

2

LE77

3

ND

203

B7

3

A634

A6

35

N28

H

100

B6

4

B14A

B6

8

CM

105

B8

4

A632

N

196

LH

195

LH

205

LH

196

LH

194

99

1

LH74

PH

G39

LH

132

G

80

7800

4

7801

0

4676

A

7800

2A

794

LP

5

PHT5

5

H84

31

1 2 3

NSS

0.0

0.1

0.2

0.3

0.4

unde

fined

C

103

O

H43

W

F9

A73

K5

5

A375

A5

56

I159

P8

38

-11

B2

33

-16

H

5

HY

K1

55

OH

40B

K4

W

22 A

B10

B3

7

T8

CO

109

O

H07

W

182B

C

I187

-2

CI5

40

Ill.1

2E

AH83

W

D

CM

37

M14

B7

0

H99

A6

54

VA26

W

117

B5

5

MO

17

R16

8

LH38

LH

39

LH51

PH

G35

M

BNA

LH

57

PHR

36

LH59

M

BST

PH

J31

PH

M57

PH

N37

PH

N73

1 2 3

IDT1 2 3

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

W

F9

A375

A5

09

A556

I1

98

B2

H5

H

Y

B164

G

N

D20

3

B10

B1

4

L317

I2

05

CI1

87-2

C

49A

I2

24

AH83

Tr

9-1-

1-6

C

11

IDT

A6

38

M14

20

7

PHN

34

PHP7

6

PHW

86

PHG

50

PHG

35

PHG

71

IB01

4

PHG

83

LH15

0

PHG

29

PHG

72

PHG

84

PHZ5

1

PHV7

8

PHK4

2

PHN

11

PHH

93

PHJ3

3

PHN

73

PHR

62

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

exPVP1960’s1940’s

1960’s1940’s exPVP

IodentNSSSS

time

line origin

SS

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

C

I31A

H

Y

B8

B10

B1

4

B96

B3

7

I205

O

H07

C

I3A

C

I187

-2

CI5

40

I224

W

D45

6

Oh3

167B

Tr

9-1-

1-6

FE

2

LE77

3

ND

203

B7

3

A634

A6

35

N28

H

100

B6

4

B14A

B6

8

CM

105

B8

4

A632

N

196

LH

195

LH

205

LH

196

LH

194

99

1

LH74

PH

G39

LH

132

G

80

7800

4

7801

0

4676

A

7800

2A

794

LP

5

PHT5

5

H84

31

1 2 3

NSS

0.0

0.1

0.2

0.3

0.4

unde

fined

C

103

O

H43

W

F9

A73

K5

5

A375

A5

56

I159

P8

38

-11

B2

33

-16

H

5

HY

K1

55

OH

40B

K4

W

22 A

B10

B3

7

T8

CO

109

O

H07

W

182B

C

I187

-2

CI5

40

Ill.1

2E

AH83

W

D

CM

37

M14

B7

0

H99

A6

54

VA26

W

117

B5

5

MO

17

R16

8

LH38

LH

39

LH51

PH

G35

M

BNA

LH

57

PHR

36

LH59

M

BST

PH

J31

PH

M57

PH

N37

PH

N73

1 2 3

IDT1 2 3

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

W

F9

A375

A5

09

A556

I1

98

B2

H5

H

Y

B164

G

N

D20

3

B10

B1

4

L317

I2

05

CI1

87-2

C

49A

I2

24

AH83

Tr

9-1-

1-6

C

11

IDT

A6

38

M14

20

7

PHN

34

PHP7

6

PHW

86

PHG

50

PHG

35

PHG

71

IB01

4

PHG

83

LH15

0

PHG

29

PHG

72

PHG

84

PHZ5

1

PHV7

8

PHK4

2

PHN

11

PHH

93

PHJ3

3

PHN

73

PHR

62

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

exPVP1960’s1940’s

1960’s1940’s exPVP

IodentNSSSS

time

line origin

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

exPVP1960’s1940’s

SS

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

C

I31A

H

Y

B8

B10

B1

4

B96

B3

7

I205

O

H07

C

I3A

C

I187

-2

CI5

40

I224

W

D45

6

Oh3

167B

Tr

9-1-

1-6

FE

2

LE77

3

ND

203

B7

3

A634

A6

35

N28

H

100

B6

4

B14A

B6

8

CM

105

B8

4

A632

N

196

LH

195

LH

205

LH

196

LH

194

99

1

LH74

PH

G39

LH

132

G

80

7800

4

7801

0

4676

A

7800

2A

794

LP

5

PHT5

5

H84

31

1 2 3

NSS

0.0

0.1

0.2

0.3

0.4

unde

fined

C

103

O

H43

W

F9

A73

K5

5

A375

A5

56

I159

P8

38

-11

B2

33

-16

H

5

HY

K1

55

OH

40B

K4

W

22 A

B10

B3

7

T8

CO

109

O

H07

W

182B

C

I187

-2

CI5

40

Ill.1

2E

AH83

W

D

CM

37

M14

B7

0

H99

A6

54

VA26

W

117

B5

5

MO

17

R16

8

LH38

LH

39

LH51

PH

G35

M

BNA

LH

57

PHR

36

LH59

M

BST

PH

J31

PH

M57

PH

N37

PH

N73

1 2 3

IDT1 2 3

0.0

0.1

0.2

0.3

0.4

unde

fined

O

H43

W

F9

A375

A5

09

A556

I1

98

B2

H5

H

Y

B164

G

N

D20

3

B10

B1

4

L317

I2

05

CI1

87-2

C

49A

I2

24

AH83

Tr

9-1-

1-6

C

11

IDT

A6

38

M14

20

7

PHN

34

PHP7

6

PHW

86

PHG

50

PHG

35

PHG

71

IB01

4

PHG

83

LH15

0

PHG

29

PHG

72

PHG

84

PHZ5

1

PHV7

8

PHK4

2

PHN

11

PHH

93

PHJ3

3

PHN

73

PHR

62

1960’s1940’s exPVP

IodentNSSSS

time

line origin

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3

effe

ctiv

e #

anc

esto

rs

exPVP1960’s1940’s

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3

effe

ctiv

e #

anc

esto

rs

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3

dive

rsity

of a

nces

tors

exPVP1960’s1940’s

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3

effe

ctiv

e #

anc

esto

rs

1 2 30 1 2 3era

C0.

000.

100.

200.

30

BSS NSS IDT

man

hatta

n di

stan

ce

050

0015

000

2500

0

BSS NSS IDT

inv.

sim

pson

010

2030

40

Lanc

aste

r

Min

13

Rei

d

Mid

land

S. d

ent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a. c.b.

d. e.

era

1 2 3 1 2 3 1 2 3

1 2 3 1 2 3 1 2 3

dive

rsity

of a

nces

tors

exPVP1960’s1940’s

0 1 2 3

01

23

4

6 snp 5 snp 8 snp 7 snp 10 snp 9 snp 15 snp 15 snp

erara

tio o

bser

ved/

rand

omiz

ed0 10000 20000 30000 40000 50000 60000

0.0

0.1

0.2

0.3

0.4

0.5

0.6

bp

Mora

n’s

I

12

30

12

3er

a

C0.000.100.200.30

BS

SN

SS

IDT

manhattan distance

050001500025000

BS

SN

SS

IDT

inv.simpson

010203040

Lancaster

Min13

Reid

Midland

S. dent

pop

flint

0.0

0.1

0.2

0.3

0.4

0.5

a.c.

b.

d.e.

era

12

31

23

12

3

12

31

23

12

3

!

sele

ctio

nan

cest

ry

devia

tion

num

ber

ance

stor

s

0 1 2 3

0.0

0.2

0.4

0.6

0.8

1.0

era

p

!

sele

ctio

nan

cest

ry

devia

tion

num

ber

ance

stor

s

0 1 2 3

0.0

0.2

0.4

0.6

0.8

1.0

era

p

!

sele

ctio

nan

cest

ry

devia

tion

num

ber

ance

stor

s Div

ersi

ty

Genome Sequence

Selective Sweep

!

sele

ctio

nan

cest

ry

devia

tion

num

ber

ance

stor

s

!

sele

ctio

nan

cest

ry

devia

tion

num

ber

ance

stor

s

10,000 ft view: drift and diversity loss

10,000 ft view: drift and diversity loss

• increasingly small, homogeneous germplasm making up ancestry of modern lines

10,000 ft view: drift and diversity loss

• increasingly small, homogeneous germplasm making up ancestry of modern lines

• changing ancestry not selection (sweeps) drives diversity across all heterotic groups

10,000 ft view: drift and diversity loss

• increasingly small, homogeneous germplasm making up ancestry of modern lines

• changing ancestry not selection (sweeps) drives diversity across all heterotic groups

• no evidence that popular lines have more good alleles

Genetic change within a single program: BSSS/BSCB1

Gerke et al. 2015 Genetics

Genetic change within a single program: BSSS/BSCB1

Gerke et al. 2015 Genetics

BSSS1

BSCB11

BSSS1

BSCB11

BSSS1

BSCB11

S1

S1

BSSS1

BSCB11

yield trials

S1

S1

BSSS1

BSCB11

yield trials

S1

S1

Ne~20

BSSS1

BSCB11

yield trials

S1

S1

BSSS2

BSCB2

Ne~20

Morell, Buckler, and Ross-Ibarra. Nat. Rev. Genetics. 2012

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Morell, Buckler, and Ross-Ibarra. Nat. Rev. Genetics. 2012

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Morell, Buckler, and Ross-Ibarra. Nat. Rev. Genetics. 2012

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Box 1 | Genetic load

Genetic load refers to the reduction in fitness caused by suboptimal genotypes in a population121. Genetic load can arise in a number of ways, including directional selection, recombination or mutation. Mutational load — the presence of deleterious mutations segregating in a population — is of particular interest for crop genomics. Deleterious mutations are most readily detected in protein-coding genes and can take several forms, including premature stop codons, splice site variants or insertions and deletions (indels) that result in the loss or impairment of protein function. These types of mutations are frequently associated with Mendelian disorders in humans, providing direct evidence that loss-of-function changes tend to be deleterious, particularly when homozygous122. Although most nonsynonymous mutations in plants are strongly deleterious, a sizable proportion are only slightly so, and these mutations may segregate at appreciable frequencies123.

Unambiguously deleterious mutations are fairly common in crop genomes17,54,124. Statistical analysis of homologous sequence from multiple genomes can identify amino acid changes that are likely to be disadvantageous (for example, REF. 125), but these comparative analyses benefit from transcriptomic data, as transcript variation among individuals may render some putatively deleterious mutations inconsequential120. Part a of the figure shows a hypothetical alignment of coding sequence from multiple grass species. The conserved nature of the histidine amino acid across species suggests that the nonsynonymous change (indicated by the red ‘G’) observed in maize is likely to be deleterious. Synonymous changes are shown in black.

Selection against deleterious mutations is hindered by Hill–Robertson effects — because of linkage, selection can only act on the net effect of both beneficial and deleterious mutations. Deleterious mutations should thus be enriched in regions of the genome in which recombination is suppressed and around the targets of strong positive selection126,127. Although neither prediction has yet been explicitly demonstrated in crops, patterns of residual heterozygosity in the maize genome support the first prediction14, and evidence from humans128 bears out the second. Whereas inbreeding can act to purge deleterious mutations129,130, drift can increase the frequency of deleterious mutations in small populations131,132. Drift is a stochastic process, and unique sets of deleterious alleles would be expected to increase in frequency in different breeding populations (for example, REF. 124). This is illustrated in part b of the figure, in which two nonsynonymous mutations (indicated by the red ‘A’s) in the ancestral population increase in frequency in two derived populations. Because drift operates independently in isolated populations, different breeding programs are likely to have a number of distinct, high-frequency deleterious mutations. Given that most deleterious mutations are at least partially recessive, crosses between lines from different breeding populations should exhibit complementation at these loci, explaining, at least in part, the widespread observation of heterosis.

Nature Reviews | Genetics

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A A C G A C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G A C T T C . . .

. . . A A C G C C T T C . . .

. . . A G A G G A C T C . . .

. . . C G A G G C C T C . . .

. . . A G G G G A C T C . . .

. . . A G A G G A C T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T C . . .

. . . A G A A G A C T A . . .

. . . A G A A G A C T C . . .

. . . A G A G G A C T C . . .

. . . A G A A G A C T C . . .

Derived population 1 Derived population 2

. . . A A T G C C T T C . . .

. . . A A C G C C T T T . . .

. . . A A C G C C T T T . . .

. . . A G G G G A C T C . . .

. . . A G A A G A C T C . . .

Gene 1

a

Gene 2

Gene 2 Gene 1 Gene 2

. . . A A C G A T C T C . . .

HisAsn Leu

AspAsn Leu

. . . A A T C A T C T C . . .

. . . A A C C A C C T C . . .

. . . A A C C A C C T T . . .

. . . A A T G C G T T C . . .

. . . A A C G C G T T C . . .

Ancestral populationb

Rice

Brachypodium

Sorghum

Maize

Gene 1

REVIEWS

NATURE REVIEWS | GENETICS ADVANCE ONLINE PUBLICATION | 5

© 2012 Macmillan Publishers Limited. All rights reserved

Complementation & Hybrid Vigor

Genetic change within a single program: BSSS/BSCB1

Genetic change within a single program: BSSS/BSCB1

• genetic drift explains most change in diversity

Genetic change within a single program: BSSS/BSCB1

• genetic drift explains most change in diversity

• little overlap in selected regions

Genetic change within a single program: BSSS/BSCB1

• genetic drift explains most change in diversity

• little overlap in selected regions

• complementation of deleterious alleles rather than overdominance likely basis of heterosis

How important are deleterious variants?

Mezmouk & Ross-Ibarra G3 2014

similar AAlikely neutral

similar AAlikely neutral

different AAlikely deleterious

similar AAlikely neutral

different AAlikely deleterious

nss

ts

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

NSS

TRO

PICA

LDeleterious allele frequency

nss

ts

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

NSS

TRO

PICA

L

nss

ts0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ss

nss

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1SS

NSS

Deleterious allele frequency

nss

ts

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

NSS

TRO

PICA

L

nss

ts0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

00.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

ss

nss

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1SS

NSS

Deleterious allele frequency

23

45

67

89

chr10

−log10(p)

20 40 60 80 100 120 140

●●● ●

●●

Chromosome 1

Prop

ortio

n no

nsyn

omyo

us

0.0

0.4

0.8

7 42 77 119 168 217 266

Chromosome 1

Prop

ortio

n no

nsyn

omyo

us

0.0

0.4

0.8

7 42 77 119 168 217 266

Chromosome 1

Prop

ortio

n no

nsyn

omyo

us

0.0

0.4

0.8

7 42 77 119 168 217 266

Gore et al. 2009 ScienceLarièpe et al. 2012 Genetics

Gore et al. 2009 ScienceLarièpe et al. 2012 Genetics

How important are deleterious variants?

How important are deleterious variants?

• deleterious alleles common, usually at low frequency in at least one group

How important are deleterious variants?

• deleterious alleles common, usually at low frequency in at least one group

• all traits show enrichment of genes with deleterious alleles

How important are deleterious variants?

• deleterious alleles common, usually at low frequency in at least one group

• all traits show enrichment of genes with deleterious alleles

• complementation of deleterious alleles in low recombination regions likely important for heterosis

Experimental test of deleterious complementation

Yang et al. bioRxiv 2017

B73 Mo17 PHZ51

B73

Mo17

PHZ51

B73 Mo17 PHZ51

B73

Mo17

PHZ51

B73 Mo17 PHZ51

B73

Mo17

PHZ51

Flowering Time

Height

Yield

GERP = Neutral rate - Estimated rate

High GERP (high function)

Low GERP (low function)

●●●

●●●

TW DTP

DTS AS

I

PHT

EHT

GY

0

50

100

150

200

BPH

(100

%)

●●

●●●●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●

●●●

●●

●●●●

●●

●●●

●●

●●

●●●

●●

●●●●

●●

●●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●●●

●●

●●●

●●●●

●●

●●●

●●

●●●

●●

●●

●●●

●●

●●●●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●

●●

●●

●●●

●●

●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●●●

●●●

●●

●●

●●●

●●

●●

●●

●●

●●●●

●●

●●●

●●●●●●●●●

●●

●●●●●●

●●

●●●

●●●●●●●

●●

●●●

●●

●●

●●

●●●●

●●

●●

●●

●●●●

●●●●

●●

●●●●●●

●●●

●●

●●●●●

●●●

●●●

●●●●

●●●

●●

●●●●

●●

●●●●●●

●●

●●●●●●●

●●

●●●●●

●●

●●

●●

●●

●●

●●

●●●

●●

●●

●●●

●●●●●

●●

●●●

●●

●●●●●●

●●●

●●

●●●●

●●

●●

●●●

●●●

●●

●●●

●●

●●●●●

●●●

●●●

●●

●●●●

●●

●●

●●

●●●

●●●

●●

●●●

●●●

●●

●●

●●●

●●

●●●●●

●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●

●●

●●

●●●

●●

●●

●●●●

●●

●●

●●

●●

●●●●

●●

●●●

●●●●

●●

●●

●●

●●

●●

●●●

●●

●●

●●

●●●●

●●

●●

●●

●●

●●

●●●

●●●●●●

●●

●●

●●

●●●●

●●

●●

●●●

●●●

●●

●●

●●●●

●●

●●●

●●

●●

●●●●●

●●

●●

●●

●●

●●

●●

●●●●

●●

●●

●●●

●●●●●●

●●

●●

●●

●●

●●●●●

●●●

●●

●●

●●

●●

●●

●●

●●

●●●

●●

●●

●●

●●●

●●

●●

●●

●●●●●●●

●●●

●●●●

●●

●●

●●●●●●

●●

●●

●●●

●●●

●●●

●●

−10 −5 0 5

0.0

0.2

0.4

0.6

GERP Score

Del

eter

ious

Alle

le F

requ

ency

LR MZ LR MZ LR MZ

0.08

0.12

0.16

0.20

Del

eter

ious

Loa

d pe

r bp

0.6

0.8

1.0

1.2

Quantiles of cM/Mb

GER

P Sc

ore

25 50 75 100

a b

c d

All Sites

Fixed Segregating

aj, additive effect of the jth GERP-SNP; Xij, 0-1-2 coding of jth GERP-SNP on ith hybrid;

dj, dominance effect of the jth GERP-SNP; Wij, 0-1-0 coding of jth GERP-SNP on ith hybrid.

Heterosis increasing

Height YieldFlowering Time

aj, additive effect of the jth GERP-SNP; Xij, 0-1-2 coding of jth GERP-SNP on ith hybrid;

dj, dominance effect of the jth GERP-SNP; Wij, 0-1-0 coding of jth GERP-SNP on ith hybrid.

Heterosis increasing

Height YieldFlowering Time

aj, additive effect of the jth GERP-SNP; Xij, 0-1-2 coding of jth GERP-SNP on ith hybrid;

dj, dominance effect of the jth GERP-SNP; Wij, 0-1-0 coding of jth GERP-SNP on ith hybrid.

k > 1 Overdominance

k = 1 Dominance

k = -1 Recessive

k < -1 Underdominance

k = 0 Additive

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Add

itive

Effe

ct

0.00

0.01

0.02

0.03

0.0 0.5 1.0 1.5 2.0GERP Score

Dom

inan

t Effe

ct

0.0

0.1

0.2

0.3

0.4

0.0 0.5 1.0 1.5 2.0GERP Score

Deg

ree

of D

omia

nce

(k)

TW DTP DTS ASI PHT EHT GY

−1

0

1

2

Deg

ree

of D

omin

ance

(k)

TraitsTWDTPDTSASIPHTEHTGY3.

5e−0

64.

0e−0

64.

5e−0

65.

0e−0

6

0.1 0.2 0.3 0.4 0.5

Allele Frequency

Varia

nce

Expl

aine

d

ba

c d e

GenotypeGERP Scores

*

* * *

YieldFlowering Height

GERP

YieldFlowering Height

random

Experimental test of deleterious complementation

• yield shows more dominance than other traits

• how deleterious an allele is matters for yield

• deleterious alleles are recessive (for yield)

• modeling complementation improves prediction of hybrid yield and heterosis 5-10%

Heterosis yield

Duvick 2005 Maydica

Hybrid yield

Inbred yield

Unasked for opinions on heterotic groups from a guy who knows nothing

about breeding

• The Good:

• intellectual & genetic control of germplasm

• hybrid vigor (it’s not all dominance)

• The Bad:

• diversity loss

• inefficient selection

Unasked for opinions on heterotic groups from a guy who knows nothing

about breeding

• Option 1:

• heterotic groups, but large Ne and genotype to enrich for recombination

• Option 2:

• mass (genomic) selection on randomly mated populations

Joost van Heerwaarden

Justin Gerke

Sofiane Mezmouk

Jinliang Yang

Wageningen University

Dupont Pioneer KWS U. Nebraska Lincoln

top related