Genetics for Epidemiologists Lecture 2: Measurement of Genetic Exposures

1
Genetics for EpidemiologistsLecture 2
Measurement of Genetic Exposures
National Human Genome Research Institute
U.S. Department of Health and Human
Services National Institutes of Health National
Human Genome Research Institute
National Institutes of Health
Teri A. Manolio, M.D., Ph.D.Director, Office of
Population Genomics and Senior Advisor to the
Director, NHGRI, for Population Genomics
U.S. Department of Health and Human Services
2
Topics to be Covered

• Measuring genetic variation
• Blood group markers
• Restriction-fragment length polymorphisms
• Variable number of tandem repeats (VNTRs,
  minisatellites and microsatellites)
• Single nucleotide polymorphisms (SNPs)
• Linkage disequilibrium (LD)
• Familial resemblance and family history

3
Larson, G. The Complete Far Side. 2003.
4
Measuring Genetic Variation Blood Group and
Enzymatic Markers

• RBC COMT activity measured in 5 large families
  with hypertension (total 518 individuals)
• Associations tested with 25 genetic markers ABO,
  Rh, K, MNS, P, Fy, Jk, PGD, ADA, ACP1, PGM1, HBB,
  GPT, C3, HPA, TF, GC, OR, GM, KM, BF, ESD, GLO1,
  Le
• Lod score of 1.27 and estimated recombination
  fraction of 0.1 found for phosphogluconate
  dehydrogenase (PGD)

Am J Med Genet 1984 19525-32.
5
Restriction Fragment Length Polymorphisms (RFLPs)

• Define polymorphic marker loci that can be
  detected as differences in length of DNA
  fragments after digestion with DNA
  sequence-specific endonucleases
• Establish linkage relationships using pedigree
  analysis

Am J Hum Genet 1980 32314-331.
6
Restriction Fragment Length Polymorphisms (RFLPs)

• Since the RFLPs are being used simply as genetic
  markers, any trait segregating in a pedigree can
  be mapped. Such a procedure would not require
  any knowledge of the biochemical nature of the
  trait or of the nature of the alterations in the
  DNA responsible for the trait.

Am J Hum Genet 1980 32314-331.
7
RFLPs Used to Map Neurofibromatosis
  
  

• Linkage analysis of 15 Utah kindreds showed
  that a gene responsible for von Recklinghausen
  neurofibromatosis (NF) is located near the
  centromere on chromosome 17

Science 1987 2361100-1102.
8
RFLPs Used to Map Neurofibromatosis
  
  

• Cosegration of NF with the A2 (1.9 kb) allele
  and not A1 (2.4kb) in each of four affected
  offspring.

Science 1987 2361100-1102.
9
Variable Numbers of Tandem Repeats (VNTRs)
Minisatellites

• Repetition in tandem of a short (6- to 100-bp)
  motif spanning 0.5 kb to several kb
• Opened the way to DNA fingerprinting for
  individual identification
• Provided the first highly polymorphic,
  multiallelic markers for linkage studies
• Associated with many interesting features of
  human genome biology and evolution
• Well-known minisatellite is 5.5kb, kringle IV
  repeat in apolipoprotein(a) and plasminogen

Vernaud G and Denoued F, Genome Res 2000
10899-907.
10
Kringle-IV Encoding Sequences of Human apo(a)
cDNA ApoA1 Alleles
Lackner et al, Hum Mol Genet 1993 2933-40.
11
Correlations of ApoA Molecular Weight with Lp(a)
Levels and Number of Kringle-IV Repeats
Gavish et al, J Clin Invest 1989 842021-27.
12
Simple Sequence Repeats (also VNTRs)
Microsatellites
Repetition in tandem of a short (2- to 6-bp)
motif from 5-5,000 times

• Most are di-, tri-, and tetra-nucleotide repeats
  repeated 20-50 times
• Most are highly polymorphic making them
  enormously useful for mapping and linkage
• Marshfield and similar maps placed 400
  microsatellites across genome, provided primers
  for analysis
• Could be highly automated NHLBI and CIDR
  large-scale genotyping services

13
Multipoint LOD Scores for Long-term SBP and DBP
on Chromosome 17
Levy et al, Hypertension 200036477-483.
14
Larson, G. The Complete Far Side. 2003.
15
Single Nucleotide Polymorphisms (SNPs)
GAAATAATTAATGTTTTCCTTCCTTCTCCTATTTTGTCCTTTACTTCAA
TTTATTTATTTATTATTAATATTATTATTTTTTGAGACGGAGTTTC/ACT
CTTGTTGCCAACCTGGAGTGCAGTGGCGTGATCTCAGCTCACTGCACACT
CCGCTTTCCTGGTTTCAAGCGATTCTCCTGCCTCAGCCTCCTGAGTAGCT
GGGACTACAGTCACACACCACCACGCCCGGCTAATTTTTGTATTTTTAGT
AGAGTTGGGGTTTCACCATGTTGGCCAGACTGGTCTCGAACTCCTGACCT
TGTGATCCGCCAGCCTCTGCCTCCCAAAGAGCTGGGATTACAGGCGTGAG
CCACCGCGCTCGGCCCTTTGCATCAATTTCTACAGCTTGTTTTCTTTGCC
TGGACTTTACAAGTCTTACCTTGTTCTGCC/TTCAGATATTTGTGTGGTC
TCATTCTGGTGTGCCAGTAGCTAAAAATCCATGATTTGCTCTCATCCCAC
TCCTGTTGTTCATCTCCTCTTATCTGGGGTCACA/CTATCTCTTCGTGAT
TGCATTCTGATCCCCAGTACTTAGCATGTGCGTAACAACTCTGCCTCTGC
TTTCCCAGGCTGTTGATGGGGTGCTGTTCATGCCTCAGAAAAATGCATTG
TAAGTTAAATTATTAAAGATTTTAAATATAGGAAAAAAGTAAGCAAACAT
AAGGAACAAAAAGGAAAGAACATGTATTCTAATCCATTATTTATTATACA
ATTAAGAAATTTGGAAACTTTAGATTACACTGCTTTTAGAGATGGAGATG
TAGTAAGTCTTTTACTCTTTACAAAATACATGTGTTAGCAATTTTGGGAA
GAATAGTAACTCACCCGAACAGTG/TAATGTGAATATGTCACTTACTAGA
GGAAAGAAGGCACTTGAAAAACATCTCTAAACCGTATAAAAACAATTACA
TCATAATGATGAAAACCCAAGGAATTTTTTTAGAAAACATTACCAGGGCT
AATAACAAAGTAGAGCCACATGTCATTTATCTTCCCTTTGTGTCTGTGTG
AGAATTCTAGAGTTATATTTGTACATAGCATGGAAAAATGAGAGGCTAGT
TTATCAACTAGTTCATTTTTAAAAGTCTAACACATCCTAGGTATAGGTGA
ACTGTCCTCCTGCCAATGTATTGCACATTTGTGCCCAGATCCAGCATAGG
GTATGTTTGCCATTTACAAACGTTTATGTCTTAAGAGAGGAAATATGAAG
AGCAAAACAGTGCATGCTGGAGAGAGAAAGCTGATACAAATATAAAT/GA
AACAATAATTGGAAAAATTGAGAAACTACTCATTTTCTAAATTACTCATG
TATTTTCCTAGAATTTAAGTCTTTTAATTTTTGATAAATCCCAATGTGAG
ACAAGATAAGTATTAGTGATGGTATGAGTAATTAATATCTGTTATATAAT
ATTCATTTTCATAGTGGAAGAAATAAAATAAAGGTTGTGATGATTGTTGA
TTATTTTTTCTAGAGGGGTTGTCAGGGAAAGAAATTGCTTTTT SNPs
1 / 300 bases 10 million across genome

16
Mapping the Relationships Among SNPs
Christensen and Murray, N Engl J Med 2007
3561094-1097.
17
Chromosome 9p21 Region Associated with MI
Samani N et al, N Engl J Med 2007 357443-453.
18
Distances Among East Coast Cities
Boston Provi-dence New York Phila-delphia Balti-more
Providence 59
New York 210 152
Philadelphia 320 237 86
Baltimore 430 325 173 87
Washington 450 358 206 120 34
19
Distances Among East Coast Cities
Boston Provi-dence New York Phila-delphia Balti-more
Providence 59
New York 210 152
Philadelphia 320 237 86
Baltimore 430 325 173 87
Washington 450 358 206 120 34
lt 100 101-200 201-300 301-400 gt 400
20
Distances Among East Coast Cities
Boston Provi-dence New York Phila-delphia Balti-more
Providence 59
New York 210 152
Philadelphia 320 237 86
Baltimore 430 325 173 87
Washington 450 358 206 120 34
lt 100 101-200 201-300 301-400 gt 400
21
Distances Among East Coast Cities
22
Distances Among East Coast Cities
Boston Provi-dence New York Phila-delphia Balti-more Wash-ington
23
One Tag SNP May Serve as Proxy for Many


Block 1
Block 2
SNP4 ?
SNP3 ?
SNP5 ?
SNP6 ?
SNP7 ?
SNP8 ?
SNP2 ?
SNP1 ?

• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CGGATTGCTGCATGGATCGCATCTGTAAGCAC
• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CAGATCGCTGGATGAATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAC  

24
One Tag SNP May Serve as Proxy for Many


Block 1
Block 2
SNP4 ?
SNP3 ?
SNP5 ?
SNP6 ?
SNP7 ?
SNP8 ?
SNP2 ?
SNP1 ?

• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CGGATTGCTGCATGGATCGCATCTGTAAGCAC
• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CAGATCGCTGGATGAATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAC

25
One Tag SNP May Serve as Proxy for Many


Block 1
Block 2
SNP3 ?
SNP5 ?
SNP6 ?
SNP7 ?
SNP8 ?

• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CGGATTGCTGCATGGATCGCATCTGTAAGCAC
• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CAGATCGCTGGATGAATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAC

26
One Tag SNP May Serve as Proxy for Many


Block 1
Block 2
SNP3 ?
SNP6 ?
SNP8 ?

• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CGGATTGCTGCATGGATCGCATCTGTAAGCAC
• CAGATCGCTGGATGAATCGCATCTGTAAGCAT
• CAGATCGCTGGATGAATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAT
• CGGATTGCTGCATGGATCCCATCAGTACGCAC

27
One Tag SNP May Serve as Proxy for Many
Block 1
Block 2
Frequency
Singleton

• GTT 35
• CTC 30
• GTT 10
• GAT 8
• CAT 7
• CAC 6 
• other haplotypes 4

28
Pair-Wise Linkage Disequilibrium (LD) Measures
Name Symbol Definition
"Lewontin's D" D pABpab pAbpaB
"D prime" D' D / max (D)
Correlation ("r-squared") r2 D2 / pApapBpb
For a discussion and comparison of these LD
measures, see Devlin B, Risch N, Genomics 1995
29311-22.
Courtesy K. Jacobs, NCI
29
Two Measures of LD D' and r2

• D' varies from 0 (complete equilibrium) to 1
  (complete disequilibrium)
• When D' 0, typing one SNP provides no
  information on the other SNP
• D' does not adequately account for allele
  frequencies r2 is correlation between SNPs, is
  preferred measure
• When r2 1, two SNPs are in perfect LD allele
  frequencies are identical for both SNPs, and
  typing one SNP provides complete information on
  the other

30
What can LD do for me?

• Knowledge of patterns of LD can be quite useful
  in the design and analysis of genetic data
• Design
• Estimation of theoretical power to detect
  associations
• Evaluation of degree of completeness of sampling
  of genetic variants
• Choice of most informative genetic variants to
  genotype
• Sample size increases by 1/r2 to achieve same
  power to detect association with SNP2 as SNP1

Courtesy K. Jacobs, NCI
31
Association Signal for Coronary Artery Disease on
Chromosome 9
Samani N et al, N Engl J Med 2007 357443-453.
32
Region of Chromosome 1 Showing Strong Association
with Inflammatory Bowel Disease
Duerr R et al. Science 2006 3141461-63.
33
LD Patterns in TCF7L2 Association Region
Grant et al, Nat Genet 2006 38320-23.
34
LD in Three HapMap Populations
International HapMap Consortium, Nature 2005
4371299-1320.
35
A HapMap for More Efficient Association Studies
Goals

• Use just the density of SNPs needed to find
  associations between SNPs and diseases
• Do not miss chromosomal regions with disease
  association
• Produce a tool to assist in finding genes
  affecting health and disease
• Ancestral populations differ in their degree of
  LD recent African ancestry populations are older
  and have shorter stretches of LD, need more SNPs
  for complete genome coverage

36
SNPs as Gateway to Genome-Wide Association (GWA)
Studies

• SNPs much more numerous than other markers and
  easier to assay
• Genome-wide studies attempt to capture majority
  of genomic variation (10M SNPs!)
• Variation inherited in groups, or blocks, so not
  all 10 million points have to be tested
• Blocks are shorter (so need to test more points)
  the less closely people are related
• SNP technology allows studies in unrelated
  persons, assuming 5kb 10kb lengths in common
  (300,000 1,000,000 markers)

37
www.hapmap.org
International HapMap Consortium, Nature 2005
4371299-1320.
38
www.hapmap.org
International HapMap Consortium, Nature 2007
449851-861.
39
Progress in Genotyping Technology
102
ABI TaqMan
ABI SNPlex
10
Cost per genotype (Cents, USD)
Illumina Golden Gate
Affymetrix MegAllele
Affymetrix 10K
Illumina Infinium/Sentrix
Perlegen
1
Affymetrix 100K/500K
Nb of SNPs
1
10
102
103
104
105
106
2001
2005
Courtesy S. Chanock, NCI
40
Continued Progress in Genotyping Technology
Affymetrix 500K
Illumina 550K
Illumina 650Y
Illumina 317K
Cost per person (USD)
July 2005
Oct 2006
Courtesy S. Gabriel, Broad/MIT
41
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study


42
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001

43
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000

44
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00

45
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00 20 billion

46
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00 20 billion
2008
47
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00 20 billion
2008 1,000,000
48
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00 20 billion
2008 1,000,000 0.05
49
Cost of a Genome-Wide Association Study in 2,000
People
Year Number of SNPs Cost/SNP Cost/Study
2001 10,000,000 1.00 20 billion
2008 1,000,000 0.05 1 million
50
Coverage ( SNPs tagged at r2 gt 0.8) of
Commercial Genotyping Platforms
HapMap population sample HapMap population sample HapMap population sample
Platform YRI CEU CHBJPT
Affymetrix GeneChip 500K 46 68 67
Affymetrix SNP Array 6.0 66 82 81
Illumina HumanHap300 33 77 63
Illumina HumanHap550 55 88 83
Illumina HumanHap650Y 66 89 84
Perlegen 600K 47 92 84
Manolio et al, J Clin Invest 2008 1181590-605.
51
Following the Polymorphism Literature

• Sometimes named for
• amino acid change (AGT M235T)
• nucleotide sequence (AGTR1 A1166C)
• promoter (AGT -6 G/A)
• restriction enzyme site (XbaI, PvuII, HindIII)
• gene product (APOEe2)
• legacy system (DRB10104)
• reference SNP (rs709932) or submitted SNP
  (ss1487247)
• Good sources for information OMIM, HUGO, dbSNP,
  UCSC Genome Browser

Courtesy S. Chanock, NCI
52
Other Genomic Technologies

• Sequencing measure variation at every point in
  gene or candidate region in dozens to hundreds of
  people to find functional variants
• Gene expression measure changes in mRNA
  (transcribed) in cases and controls or in
  response to stimulation
• Epigenetics measure DNA methylation or histone
  deacetylation that turns genes on and off

53
Sidney Harris, http//www.sciencecartoonsplus.com/
gallery.htm.
54
Summary Points Genotyping Methods

• Unbelievably rapid progress from small number of
  blood group markers to gt10M SNPs, CNVs,
  structural variants, sequence variants
• Technology will continue to change and will be
  challenge to keep up with difficult to know when
  ready to apply to population studies
• SNPs are currently the dominant technology (more
  to come in Lecture 4)
• Quality control is a major issue

55
Familial Resemblance?
http//en.wikipedia.org/wiki/ImageKennedy_bros.jp
gfile
56
Evidence for Genetic Influence on Disease or
Trait from Family Data

• Familial resemblance trait more similar among
  related than unrelated persons
• Familial clustering risk of disease in relative
  of case gt risk in relative of non-case or of
  general population (sibling relative risk,
  Risch's ?S)
• Distributions of continuous trait mixtures of
  distributions or commingling analysis

57
Sibling Relative Risk of Living to Age
90Centenarians vs. Those Dying at Age 73
Perls TT et al, Lancet 1998 3511560.
58
Large Representative Pedigree Showing 69 Patients
with Atrial Fibrillation
Arnar et al, Europ Heart J 2006 27708-12.
59
Strength of Extensive Genealogies

• Common diseases do not show Mendelian inheritance
  patterns
• Affected siblings infrequent in common diseases,
  but many patients may have more distant relatives
  with same disease

Degree of Relatives Risk Ratio 95 CI P-Value
1 1.77 1.67,1.88 lt 0.001
2 1.36 1.27,1.44 lt 0.001
3 1.18 1.14,1.23 lt 0.001
4 1.10 1.06,1.13 lt 0.001
5 1.05 1.02,1.07 lt 0.001
Arnar et al, Europ Heart J 2006 27708-12.
60
Familial Correlations

• Phenotypic resemblance among relatives estimated
  by regression of one relatives value
  (offspring), on that of another (parent)
• Yo µ ß (Ym Yf )/2 e
• Twice parent-offspring correlation is estimate of
  heritability
• If trait under genetic control, expect trait
  correlations among closer relatives to be greater
  than those among more distant relatives

61
Familial Correlations of Sex-Specific LV Mass,
Multiply-Adjusted
Relative Pair Pairs (n) Correlation Expected
Spouse 855 0.05 0
Parent-offspring 662 0.15 0.5
Sibling 1,486 0.16 0.5
Avuncular 369 0.06 0.25
after Post W et al, Hypertension 1997
301025-1028.
62
Assessing Familial and Genetic Nature of a
Phenotypic Trait Heritability

• Often designated as H, h2, or s2G /s2P
• Proportion of total inter-individual variation in
  the trait (s2P) or phenotypic variation,
  attributable to genetic variation (s2G)
• Population- and environment-specific parameter
• Its value, high or low, does not indicate role of
  genes in any specific individual
• Does allow one to predict expected degree of
  familial aggregation of a trait
• Traits with high heritability should prove
  fruitful in identifying trait-related genes

63
Genetic Basis of Familial Clustering of Plasma
ACE Activity
Relative N Mean (u/L) Major Gene Effect Major Gene Effect
Relative N Mean (u/L) Mean (u/L) Variance
Fathers 87 34.1 4.8 29
Mothers 87 30.7 4.0 29
Siblings 169 43.1 10.8 75
Cambien F, et. al., Am J Hum Genet 1988
43774-780.
64
Estimated Heritability Explained by GWA Findings
to Date
Estimated GWA s2G Estimated Total s2G Reference
Height 3 90 Weedon Nat Genet 2008
T2DM ?s 1.07 ?s 3.5 Zeggini/Scott Science 2007
CRP ? 10.5 30-50 Reiner/Ridker Nat Genet 2008
Psoriasis 9 _at_ 1.3 OR ?s 4-11 Liu PLoS Genet 2008
NHGRI GWA Catalog, www.genome.gov/GWAstudies
65
Hardy-Weinberg Equilibrium

• Occurrence of two alleles of a SNP in the same
  individual are two independent events
• Ideal conditions
• random mating - no selection (equal
  survival)
• no migration - no mutation
• no inbreeding - large population sizes
• gene frequencies equal in males and females)
• If alleles A and a of SNP rs1234 have frequencies
  p and 1-p, expected frequencies of the three
  genotypes are

Freq AA p2
Freq Aa 2p(1-p)
Freq aa (1-p)2
After G. Thomas, NCI
66
Summary Points Familial Clustering

• Indicator of possible genetic influence
• May over-estimate genetic component due to poor
  assessment and adjustment for shared environment
• Methods include twin studies, parent-offspring
  correlation, relative relative risk, variance
  explained
• Current genes for complex disease explain only
  tiny fraction of total heritability

67
Larson, G. The Complete Far Side. 2003.
68
(No Transcript)
69
Basic Definitions Loci, Genes, Alleles

• Locus Place on a chromosome where a specific
  gene or set of markers resides
• Quantitative trait locus (QTL) a genetic factor
  believed to influence a quantitative trait such
  as blood pressure, lipoprotein levels, etc.
• Gene Contiguous piece of DNA that can contain
  information to make or modify expression of
  specific protein(s)
• Allele A variant form of a DNA sequence at a
  particular locus on a chromosome
• Candidate gene Gene believed to influence
  expression of complex phenotypes due to known
  biologic properties of their products

After S. Chanock, NCI
70
Basic Definitions Parts of a Gene

• Exon a DNA sequence that usually specifies the
  sequence of amino acids in translation
• Intron an intervening DNA sequence removed from
  mRNA after transcription and thus does not encode
  protein in translation
• Splice site Junction of intron and exon
• Promoter region of DNA to which an RNA
  polymerase binds and initiates transcription -
  the promoter regulates gene expression by
  controlling the amount of mRNA transcribed
• Polymorphism Variation in the sequence of DNA
  among individuals

After S. Chanock, NCI
71
SNPs and Function We know so little

• Majority are silent
• No known functional change
• Some alter gene expression/regulation
• Promoter/enhancer/silencer
• mRNA stability
• Small RNAs
• Some alter function of gene product
• Change sequence of protein

Courtesy S. Chanock, NCI
72
SNPs within Genes

• Coding SNPs (cSNPs)
• Synonymous no change in amino acid
• previously termed silent but..
• Can alter mRNA stability
• DRD2 (Duan et al 2002)
• Can alter speed of translation and protein
  folding
• MDR1 (Gottesman et al 2007)
• Nonsynonymous changes amino acid (codon)
• conservative and radical
• Nonsense insertion of stop codon
• Frameshift (insertion/deletion) Disrupts codon
  sequence, rare but disruptive

After S. Chanock, NCI
73
SNPs Outside Genes

• Majority distributed throughout genome are
  silent (excellent as markers)
• Alter transcription
• Promoter, enhancer, silencer
• Regulate expression
• Locus control region, mRNA stability
• Most are assumed to be silent hitchhikers
• No function by predictive models or analysis

Courtesy S. Chanock, NCI
74
Sample Collection and Processing

• Obtaining samples for DNA preparation
• whole blood, buffy coat
• sputum
• buccal cells
• serum, urine
• pathology specimens
• placenta, excreta, other
• Purifying and quantifying DNA
• Transformed lymphocytes
• Whole genome amplification (WGA)
• Barcode individual DNAs (QC)

After S. Chanock, NCI