1. Office hours Wednesday 3-4pm 304A Stanley Hall

2. Fig. 11.26 Association mapping (qualitative)

3. Association scan, qualitative osteoarthritiscontrols C’s141797 G’s47433  2 test

4. Association scan, quantitative

5. Association vs. linkage

6. Unrelated individuals (usually) Related individuals

7. Association vs. linkage Unrelated individuals (usually) Related individuals Extreme of linkage study is one large family; less likely that phenotype has multiple genetic causes (locus heterogeneity).

8. Association vs. linkage Strong, easy to detect Unrelated individuals Related individuals

9. Association vs. linkage Strong, easy to detect, but rare in population Unrelated individuals Related individuals

10. Association vs. linkage Strong, easy to detect, but rare in population Unrelated individuals Related individuals (e.g. BRCA1)

11. Association vs. linkage Strong, easy to detect, but rare in population; may not be reflective of common disease. Unrelated individuals Related individuals

12. Association vs. linkage Strong, easy to detect, but rare in population; may not be reflective of common disease. Also, hard to collect family data. Unrelated individuals Related individuals

13. Association vs. linkage Strong, easy to detect, but rare in population; may not be reflective of common disease. Also, hard to collect family data. Common but weak effects Unrelated individuals Related individuals

14. Association vs. linkage Strong, easy to detect, but rare in population; may not be reflective of common disease. Also, hard to collect family data. Common but weak effects; need 1000’s of samples to detect Unrelated individuals Related individuals

15. Association vs. linkage Strong, easy to detect, but rare in population; may not be reflective of common disease. Also, hard to collect family data. Common but weak effects; need 1000’s of samples to detect. If no common cause, can fail. Unrelated individuals Related individuals

16. Another key feature of association mapping: resolution

17. Association vs. linkage many recombinations have happened since common ancestor; shared region is small; no co- inheritance between distant markers

18. Association vs. linkage small number of generations; individuals share big chunks of genome; can get co- inheritance between distant markers many recombinations have happened since common ancestor; shared region is small; no co- inheritance between distant markers

19. Association vs. linkage small number of generations; individuals share big chunks of genome; can get co- inheritance between distant markers many recombinations have happened since common ancestor; shared region is small; no co- inheritance between distant markers So you need very high density of markers to get signal in an association study, but you get very high spatial resolution.

20. Association vs. linkage small number of generations; individuals share big chunks of genome; can get co- inheritance between distant markers many recombinations have happened since common ancestor; shared region is small; no co- inheritance between distant markers In the “old days” of sparse markers, linkage analysis was the best strategy.

21. Causative variant very close -log(  2 p-value) rs377472

22. But there is a pitfall of association tests: “population structure”

23. Diabetes in Native Americans

24. (1971) Diabetes in Native Americans

25. (1971) Family studies indicate it is at least partly genetic, not environmental. Diabetes in Native Americans

26. Association mapping causal loci Typed IgG heavy chains with protein assay. Phenotypes can serve as markers too…

27. Association mapping causal loci Typed IgG heavy chains with protein assay. Phenotypes can serve as markers too… (Multiple proteins from chr 14 region: haplotype)

28. Association mapping causal loci

29. diabetescontrol Gm23270 no Gm13433284 Association mapping causal loci

30. diabetescontrol Gm23270 no Gm13433284 Association mapping causal loci

31. diabetescontrol Gm23270 no Gm13433284 Association mapping causal loci “Gm is protective against diabetes?”

32. Association mapping causal loci

33. Self-identified heritage

34. Most “full heritage” members don’t have the haplotype

35. Self-identified heritage Most “full heritage” members don’t have the haplotype The few without N.A. heritage are much more likely to have the haplotype

36. Gm haplotype is very rare in self-identified 100% Pima members.

37. Gm is a marker for Caucasian ancestry.

38. Association and admixture

40. huge almost negligible

41. Association and admixture these are all the Caucasians…

42. Association and admixture Gm doesn’t look like it has much additional protective effect if you stratify by familial origin first!

43. Association and admixture Caucasian ancestry is associated with Gm haplotype.

44. Association and admixture Caucasian ancestry is associated with Gm haplotype. Caucasian ancestry is associated with lower diabetes risk.

45. Association and admixture Caucasian ancestry is associated with Gm haplotype. Caucasian ancestry is associated with lower diabetes risk. But Gm is not associated with lower diabetes risk!

46. Association and admixture Cases Controls Controls are enriched for Caucasians

47. Association and admixture Cases Controls = = N.A. and Caucasians are different at many loci http://www.ucl.ac.uk/~ucbhjow/medicine/images/human_karyotype.gif

48. Association and admixture Cases Controls = = At any one of these loci, Caucasian-like allele will be enriched in control samples.

49. Association and admixture Cases Controls = = Don’t believe any one locus is causative!

50. Microarrays and genotyping

51. DNA microarrays Fig. 10.21

52. DNA microarrays Fig. 10.21 Post-genome era: the sequence and location of the oligos are known

53. DNA microarrays Fig. 1.13

54. Genotyping by array Fig. 11.8

55. Genotyping by array Fig. 11.10

56. Genotyping by array Fig. 11.10 oligo (human genome fragment) index

57. Genotyping by array Fig. 11.10 oligo (human genome fragment) index

58. Genotyping by array Fig. 11.10 oligo (human genome fragment) index

59. Genotyping by array Fig. 11.10 oligo (human genome fragment) index

60. Genotyping by array Fig. 11.10 middle nucleotide on chip oligo oligo (human genome fragment) index

61. Genotyping by array Fig. 11.10 middle nucleotide on chip oligo Sample DNA is labeled, allowed to hybridize

62. Genotyping by array Fig. 11.10 readout of sample middle nucleotide on chip oligo

63. Genotyping by array Fig. 11.10 readout of sample middle nucleotide on chip oligo

64. Genotyping by array Fabrication technology allows millions of oligos (each present in millions of copies) on a single slide

65. Genotyping by single-base extension Fig. 11.11

66. Genotyping by single-base extension http://www.illumina.com/downloads/InfiniumIIWorkflow.pdf

67. High density of markers necessary for association