CORSO DI EPIDEMIOLOGIA GENETICA DI BASE 11 13 ottobre 2004 Istituto Superiore di Sanit Roma STUDI DI

1
Statistical genetic methods for disease gene
identification
modified from D. Altschuler
2
Association studies

• Association between risk factor and disease risk
  factor is significantly more frequent among
  affected than among unaffected individuals
• In genetic epidemiology
• Risk factors alleles/genotypes/haplotypes

3
Association studies

• Candidate genes (functional or positional)
• Fine mapping in linkage regions
• Genome wide screen

4
Candidate gene analysis

• Direct analysis
• Association studies between disease and
  functional SNPs (causative of disease) of
  candidate gene

5
Candidate gene analysis

• Indirect analysis
• Association studies between disease and random
  SNPs within or near candidate gene
• Linkage Disequilibrium mapping

6
Association studies

• POPULATION-BASED
• Case-control studies
• FAMILY-BASED
• Cases and related controls
• Nuclear or extended pedigrees

7
Case-control studies ?2 test
Risk factor
contingency table
Test of independence ?2 ? (O-E)2 / E with
1 df
8
Case-control studies ?2 test
2x3 contingency table
Genotypes
AA Aa aa Cases nAA nAa naa N Controls
mAA mAa maa M tAA tAa taa NM
Test of independence ?2 ? (O-E)2 / E with
2 df
9
Case-control studies ?2 test
2x2 contingency table
Alleles
A a Cases nA na 2N Controls
mA ma 2M tA ta 2(NM)
Test of independence ?2 ? (O-E)2 / E with
1 df
10
Hardy-Weinberg Equilibrium

• Biallelic locus A, a genotypes AA, Aa, aa
• Allele frequencies A P(A) p
• a P(a) q
• Genotype frequencies are in HWE if
• AA P(AA) p2
• Aa P(Aa) 2pq
• aa P(aa) q2

11
Haplotypes
GENOTYPES
Locus 1
2
1
3
Locus 2
6
1
1
5
9
1
7
4
9
1
Identification of phase
6
2
9
1
7
2
1
2
1
2
7
6
1
4
1
7
1
8
1
8
1
4
Locus N
1
0
1
0
12
Motivation for haplotype-based analysis

• Increased ability to identify regions that are
  shared identical by descent among affected
  individuals, and therefore more informative
• Haplotype may be the causative composite allele
  rather than a particular nucleotide at a
  particular SNP
• Haplotype analysis is meaningful only if SNPs are
  in linkage disequilibrium

13
Haplotype determination options

• Collect and genotype family members
• Laboratory-based techniques
• Statistical estimation in unphased individuals
• Likelihood-based E-M algorithms

14
Case-control studies ?2 test
2xr contingency table
Haplotypes
1 2 3 ... r Cases n1 n2 n3 ... nr
2N Controls m1 m2 m3 ... mr
2M t1 t2 t3 ... tr 2(NM)
Test of independence ?2 ? (O-E)2 / E with
r-1 df
15
Measures of association

• Relative risk (prospective studies)

Odds ratio (retrospective studies)
16
Measures of association
Alleles A, a genotypes AA, Aa, a,a Genotype
relative risk

• risk for AA
• risk for aa
• risk for Aa
• risk for aa
• Allele relative risk
• risk for A
• risk for a

GRRAA
GRRAa
GRRa/a 1
FA
Fa 1
17
Measures of association
disease genotype - A/A n11 n12 A/a n2
1 n22 a/a n31 n32
GRRA/A ORA/A n11n32/n12n31
GRRA/a ORA/a n21n32/n22n31
disease allele - A n11 n12 a n21
n22
FA ORA n11n22/n12n21
18
Measures of association

• Genotypes
• Dominant/recessive/codominant
• - e.g. GRRAA GRRAa A dominant
• Alleles/haplotypes
• Multiplicative model GRRij Fi Fj
• HWE

19
Causes of genetic association

• Direct association

20
Causes of genetic association

• Direct association

• Indirect association due to LD

marker locus
Functional variant
LD
disease
21
Linkage disequilibrium

• Non random association between alleles at
  different loci
• Loci are in LD if alleles are present on
  haplotypes in different proportions that expected
  based on allele frequencies

22
Linkage disequilibrium

• Locus 1 alleles A and a frequencies pA and
  pa
• Locus 2 alleles B and b frequencies pB and
  pb

A
A
a
a
Possible haplotyes
B
b
B
b
D pAB - pApB ? 0
23
Measures of LD
Locus 2 B b Locus 1 A pApBD pApB-D
pA a papB-D papbD pa pB pb
D D / Dmax Dmax Min(pApb, papB)
Dgt0 r2 D2 / (pA pa pB pb)
24
Graphic representation of LD
r2
D
GOLD
25
Zondervan Cardon, 2004
26
Causes of genetic association

• Direct association

• Indirect association due to LD

• Spurious association due to confounding factors
  (e.g., population stratification)

27
Association due to population stratification
Marchini et al, 2004
28
Population stratification

• Possible solutions
• Stratify sample based on confounding variable
• Apply test correction (Genomic Control)
• Use related controls

FAMILY-BASED association studies
29
Family-based association studies
1 4 transmitted 2 3 non transmitted
?
?
1 2
3 4
control
2 3
1 4
30
TDT Transmission Disequilibrium Test
non transmitted G g G a b g c d
G/G
G/g
transmitted
G/g
TDTG (TG-NTG)2/(TGNTG) (b-c)2/(bc)
?21