Genetic Selection Tools in

the functional Genomics

Outline



Background

–

Genetic Evaluations

–

Quantitative Genetics

–

Genomics



Integrating Genetics and Genomics



Case Study: DGAT1



Tangent: Animal Identification



Crystal Ball



Conclusions

Background



Bovine Functional Genomics Laboratory (BFGL)

–

Structural and functional genomics of cattle

–

Emphasis on health and productivity

–

Bioinformatics (storage and use of genomic data)



Animal Improvement Programs Laboratory (AIPL)

–

“Traditional” genetic improvement of dairy cattle

–

Increasing emphasis on animal health and reproduction

Traditional Selection

Programs



Estimate genetic merit for animals in a
population



Select superior animals as parents of
future generations

Genetic Evaluation System



Traditional selection has been very effective

for many economically important traits



Example: Milk yield

–

Moderately heritable

–

~30 million animals evaluated 4x/yr

–

Uses ~70 million lactation records

–

Includes ~300 million test-day records

–

Genetic improvement is near theoretical

expectation

Dairy Cattle Genetics Success

-6000

-4000

-2000

0

2000

1960

1970

1980

1990

2000

Year of Birth

B

V

M

il

k

Cows

Bulls

Dairy Cattle Genetics

Industry Cooperation

Producer

DHIA/DRPC

PDCA

NAAB

AIPL

Producer

DHIA/DRPC

PDCA

NAAB

AIPL

Genomics - Introduction



Traditional dairy cattle breeding has assumed

that an infinite number of genes each with

very small effect control most traits of

interest



Logical to expect some “major” genes with

large effect; these genes are usually called

quantitative trait loci (

QTL

)



The QTL locations are unknown!



Genetic markers can provide information

about QTL

Genetic Markers



Allow inheritance of a
region of the genome to be
followed across generations



Single nucleotide
polymorphisms (

SNiP

) are

the markers of the future!



Need lots!

–

3 million in the genome

–

10,000 initial goal

Polymorphis

m

“poly”

= many

“morph”

=

form

General
population

94%

6%

Single nucleotide
polymorphism
(SNP)

Application of Genetic

Markers

1.

Identify genetic markers or polymorphisms in
genes that are associated with changes in
genetic merit

2.

Use marker assisted selection (

MAS

) or gene

assisted selection (

GAS

) to make selection

decisions before phenotypes are available

3.

Adjust genetic merit for markers or genes in
the genetic evaluation system

QTL Identification

Geneti

c

Merit

DNA

Dat

a

Compare Genetic

Merit

QTL Identification and

Marker Assisted Selection

3.5

1.7

-0.1

-2.5

-6.2

0.7

Gene Assisted

Selection

Marker or Gene Assisted

Selection



Largest benefits are for traits that:

–

have low heritability, i.e., traits where genetics
contribute a small fraction of observed variation (e.g.,
disease resistance and fertility)

–

are difficult or expensive to measure (e.g., parasite
resistance )

–

cannot be measured selection decision needs to be
made (e.g., milk yield and carcass characteristics)



Evolution in traditional selection program by
improving estimation of genetic merit

Example: DGAT1



DGAT1: diacylglycerol
acyltransferase

–

Enzyme involved in fat sythesis

–

Identified using



Genetic marker data



Model organism (mouse) gene function
information



Cattle sequence verified candidate gene

DGAT1



Two forms of the gene in cattle

–

M = high milk (low fat) form of gene

–

F = high fat (low milk) form gene



BFGL scientists decided to characterize

the gene in North American population

–

Over 3300 animals genotyped for DGAT1 SNP

–

Approximately 2900 genotypes verified and

used in these analyses

DGAT1 – Average Differences

in Daughters of Bulls

Trait

MM-

FF

Trait

MM-

FF

Milk lbs

361

Fat%

0.13

Fat lbs

16.5

Protein

%

0.02

Protein

lbs

5.0

NM$

$24

SCS

0.05

CM$

$35

PL

0.07

FM$

$4

DPR

0.21

DGAT1 Genotypic

Frequencies

Integrating Genomics

Results



Genes will likely account for a
fraction of the total genetic
variation



Cannot select solely on gene
tests!!

Integrating Genomic Data:

An Ideal Situation!

Bull PTA

Integrating Genomic Data:

The DGAT1 NM$ Situation!

Bull PTA NM$

MM

FF

Integrating Genomic Data:

The DGAT1 Fat Situation!

Bull PTA Fat

MM

FF

Integrating Genomics

Results



Combine information

–

Ideally would incorporate genomic data into
genetic evaluation system



Adjust PTA??

–

Don’t adjust well proven animals (it’s in
there!!)

–

Adjust parent average for flush mates

–

Progeny have identical parent averages

–

Adjusting other PTA is non-trivial!

Integrating Genomic Data:

Another view of DGAT1 NM$!

Bull PTA NM$

MM

FF

And it Really Works!



Recent German study evaluated impact on
adjusting historic parent averages (

PA

) for DGAT1

and evaluated impact of predictability of future
evaluations



Correlations of original PA with eventual PTA for milk
were 45%



Correlations of adjusted PA with eventual PTA for
milk were 55% (

10% gain

)



Incorporation of genomic data will result in
increased stability of evaluations

Genetic Evaluations -

Limitations



Slow!

–

Progeny testing for production traits take 3 to 4

years from insemination

–

A bull will be at least 5 years old before his

first evaluation is available



Expensive!

–

Progeny testing costs $25,000 per bull

–

Only 1 in 8 to 10 bulls graduate from progeny

test

–

At least $200,000 invested in each active bull!!

Genetic Evaluations:

Genomics Enhancements



Faster

–

Use of gene and marker tests allow
preliminary selection decisions beyond
parent average before performance and
progeny test data are available



Cheaper

–

Improved selection decisions should result
in higher graduation rates or enhanced
genetic improvement

How do we get there



Increase number of genetic markers



Continue QTL discovery for MAS/GAS



Better characterize the genome

–

Compare genome to well characterized
human and mouse genome

Bovine Genome Sequence

Bovine Genome Sequence



Inbred Hereford is primary animal being
sequenced



Genome size is similar to humans



Sequencing about half completed



First assembly released yesterday!!

–

2.3 of 2.8 billion base pairs

–

84% coverage

L1 Dominette 01449

Bovine Genome Sequence



Six breeds selected for

low level sequencing



Holstein and Jersey cows

represent dairy breeds



Useful for SNP marker

development



Expect 3 million SNPs in

the genome



Preliminary goal is to

characterize 10,000

Wa-Del RC Blckstr Martha-ET

Mason Berretta Jenetta

Genomic Tools for

Parentage Verification



Low-cost high-throughput SNP marker tests would

facilitate parentage verification and traceability



$10 to $20 per sample seems to be a common break

point



Progeny test herds would likely be early adopters

–

Support from studs?



Results in increased stability on first proofs?

–

Nearly impossible to make mistake on parentage

–

Punished on second crop proofs?



With widespread implementation

–

Increase effective heritability

–

Decrease evaluation variability

–

Enhanced genetic improvement

Crystal Ball (Wishful

Thinking?)



Large number of validated genetic tests

available



Large amounts of marker and gene

data publicly available



Genomic data incorporated into genetic

evaluation



Management decisions facilitated by

genomics data

Considerations in Genomic

Tests



How big is the effect?

–

Traits of interest, economic index (NM$, TPI, PTI)

–

How many genetic standard deviation units?



Has this been validated by a sufficiently large
independent study?



What correlated response is expected & observed?



What are allele frequencies?



What is the value of this test?

–

not simple to answer

Conclusions



Genomics is enhancing genetic improvement



DGAT1 has large impacts on milk, fat,

protein, SCS



Genetic tests need to be weighted

appropriately for optimal selection decisions



Genomic tools will be extremely powerful for

parentage verification and traceability

–

Could impact genetic evaluations