topics in statistical genetics
TRANSCRIPT
![Page 1: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/1.jpg)
1
Topics in Statistical Genetics
Karl W Broman
Department of BiostatisticsJohns Hopkins School of Public Health
![Page 2: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/2.jpg)
2
• QTLs in experimental crosses
• CEPH data– Genetic maps
– Recombinational variation
– Autozygosity
– Crossover interference
• Genotyping/pedigree errors
• Dog genetic maps
• Various disease projects
Past
8 families (~ 130 people)>8000 STRPs{
![Page 3: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/3.jpg)
3
Traditional genetics: binary traitse.g. disease or no disease
Quantitative traitse.g. yield of tomato crop
# bristles on a fly’s abdomen
– Many genes– Environmental variation
Why?– Biochemical basis of trait– Selection experiments– Evolution
Goal: find (some of) the genes
![Page 4: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/4.jpg)
4
Backcross experiment
![Page 5: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/5.jpg)
5
Genetic markers:STRPs or microsatellites
![Page 6: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/6.jpg)
6
DataPhenotypes (trait values)
yi = phenotype for individual i
Marker genotypes
xij = 1/0 if i is HL/LL at marker j
Genetic map
Locations of markers
ModelsRecombination: No interference
Phenotype/genotype connectiony = µ + ßj zj +
~ Normal(0, 2)
![Page 7: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/7.jpg)
7
Problem
100 to 1000 backcross progeny
100 to 400 markers
y = µ + ßj xj +
Errors:
• Miss important loci
• Include extraneous loci
Find the x’ s with ßj ≠ 0
![Page 8: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/8.jpg)
8
Discussion
• Identifying QTLs is model selection
• Simulation studies are necessary– compare procedures– understand a procedure’s performance
• Different situations will require different procedures
![Page 9: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/9.jpg)
9
Human data
• www.marshmed.org/genetics
• 8 CEPH families– three generations– 11 to 15 progeny– 92 meioses
• ~8,000 STRP markers– 90 ± 7 % typed
• Average spacing– female: 0.6 ± 1.2 cM– male: 0.4 ± 1.0 cM– sex-ave: 0.5 ± 0.9 cM
• Data cleaning
– Removed 764/964,425 (~0.08%) genotypes resulting in tight double recombinants
![Page 10: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/10.jpg)
10
Meiosis
![Page 11: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/11.jpg)
11
• Strand choice
→ Chromatid interference
• Spacing
→ Chiasma (crossover) interference
Interference
![Page 12: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/12.jpg)
12
distance (cM) = average # crossoversin 100 meiotic products
perMorgan
2 chiasmata on 4-strand bundle1 crossover on meiotic product{
recombinationfraction
geneticdistance
as a function of
Haldane r(d) = ½ [ 1 – exp(-2d) ]
Kosambi r(d) = ½ tanh(2d)
![Page 13: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/13.jpg)
130 100 200 300
![Page 14: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/14.jpg)
14
15 20 25 30 35
Total Number of Crossovers
102
884
1331
1332
1347
1362
1413
1416
CE
PH
Fam
ily
Crossovers in Father
![Page 15: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/15.jpg)
15
20 25 30 35 40 45 50 55 60
Total Number of Crossovers
102
884
1331
1332
1347
1362
1413
1416
CE
PH
Fam
ily
Crossovers in Mother
![Page 16: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/16.jpg)
16
0 50 100 150 200 250 300
0
1
2
3
4
5
Chromosome 1
Sex-averaged Location
Fem
ale/
Mal
e D
ista
nce
Rat
io
![Page 17: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/17.jpg)
17
0 50 100 150 200
0
2
4
6
Chromosome 4
Sex-averaged Location
Fem
ale/
Mal
e D
ista
nce
Rat
io
![Page 18: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/18.jpg)
18
14m
13m
12m
11m
10m
9m
8m
7m
6m
5m
4m
3m
0 100 200 300 400
Chromosome 6Family 884
![Page 19: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/19.jpg)
19
Individual Number of regions
Length (cM)
Total length (cM)
884-01 (father) 10 2 – 23 116
884-02 (mother) 11 5 – 40 165
884: grandparents 2 7 – 10
7 1 – 29
14 58 – 120
884: 12 progeny 5 – 18 2 – 36 62 – 204
102: 14 progeny 4 – 13 1 – 39 85 – 254
1331-12 (GP) 2 1 – 6 7
1416-14 (GP) 4 6 – 29 76
35/100 of the others 1 – 2 0 – 7
CEPH: homozygous regions
![Page 20: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/20.jpg)
20
• Finish the past
• Yellowstone wolves
• Sib pairs where parents are unavailable
• Various disease projects
Present
![Page 21: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/21.jpg)
21
130 / 134 134 / 138
130 / 138 134 / 139
![Page 22: Topics in Statistical Genetics](https://reader036.vdocument.in/reader036/viewer/2022071602/613d5494736caf36b75c0d0e/html5/thumbnails/22.jpg)
22
• Large genetic/epidemiological studies– Many people
– Many phenotypes
– Many environmental covariates
• Highly computational statistical genetic methods
• Various disease projects
Future