1. SNP Selection University of Louisville Center for Genetics and Molecular Medicine January 10, 2008 Dana Crawford, PhD Vanderbilt University Center for Human Genetics Research

2. Outline of Tutorial Concepts of tagSNPs LD and haplotype definitions Haplotype blocks and definitions Tools to identify tagSNPs

3. Why Do We Need tagSNPs? Whole Genome: 15,000,000 SNPs 6,000,000 SNPs > 5% MAF Too Many SNPs to Genotype! Ex: E2F2 Average Gene: 26.5 kb 130 SNPs 44 SNPs ≥5% MAF

4. SNP Genotypes Are Correlated (aka linkage disequilibrium) “the nonindependence of alleles at different sites.” Pritchard and Przeworski 2001 Genotype at one site can predict genotype at another site Proportion of genotypes are correlated

5. Measuring Pair-wise SNP Correlations SNP genotype correlation described by linkage disequilibrium (LD) Pair-wise measures of LD: D´ and r 2 D = p AB - p A p B ; D´ = D/D max Recombination r 2 = D 2 f(A 1 )f(A 2 )f(B 1 )f(B 2 ) Power

6. r 2 is inversely related to power (“effective sample size”) 1/r 2 1,000 cases1,250 cases 1,000 controls r 2 =1.01,250 controlsr 2 = 0.80 D´ is related to recombination history D´ = 1no recombination D´ < 1historical recombination LD Statistics: Practical Uses

7. Where to Find Population LD Statistics For your gene or region of interest, search HapMapwww.hapmap.org Perlegengenome.perlegen.com SeattleSNPs PGApga.gs.washington.edu NIEHS SNPsegp.gs.washington.edu

8. Where to Find Population LD Statistics For your gene or region of interest, search HapMapwww.hapmap.org Perlegengenome.perlegen.com SeattleSNPs PGApga.gs.washington.edu NIEHS SNPsegp.gs.washington.edu

9. Visualizing Pair-wise LD

12. Where to Find Population LD Statistics For your gene or region of interest, search HapMapwww.hapmap.org Perlegengenome.perlegen.com SeattleSNPs PGApga.gs.washington.edu NIEHS SNPsegp.gs.washington.edu Genome Variation Server

13. Visualizing Pair-wise LD

22. Multi-SNP Genotype Correlations (aka Haplotypes) “…a unique combination of genetic markers present in a chromosome.” pg 57 in Hartl & Clark, 1997

23. Constructing Haplotypes C TA GC TA G T TG GT TG G C CA GC CA G C/T, A/G C/C, A/G T/T, G/G C/T, A/A C/C, A/G Collect pedigreesSomatic cell hybrids Human Rodent Hybrid SNP 1 SNP 2 C/TA/G Allele-specific PCR

24. Constructing Haplotypes Examples of Haplotype Inference Software: EM Algorithm Haploview http://www.broad.mit.edu/mpg/haploview/index.php Arlequin http://lgb.unige.ch/arlequin/ PHASE v2.1 http://www.stat.washington.edu/stephens/software.html HAPLOTYPER http://www.people.fas.harvard.edu/~junliu/Haplo/docMain.htm

25. Haplotypes in NIEHS SNPs >625 genes re-sequenced Cell cycle, DNA repair/replication, apoptosis 2 DNA panels 1: Polymorphism Discovery Resource (PDR90) 2: Europeans, Africans, Hispanics, and Asians PHASEv2.0 results posted on website Interactive tool (VH1) to visualize and sort haplotypes http://egp.gs.washington.edu

26. Haplotypes in NIEHS SNPs

37. Haplotypes in NIEHS SNPs

38. r 2 is inversely related to power (“effective sample size”) 1/r 2 1,000 cases1,250 cases 1,000 controls r 2 =1.01,250 controlsr 2 = 0.80 D´ is related to recombination history D´ = 1no recombination D´ < 1historical recombination Example: Tagger and LDSelect Example: Haplotype “blocks” Using LD and Haplotypes to Pick tagSNPs

39. r 2 is inversely related to power (“effective sample size”) 1/r 2 1,000 cases1,250 cases 1,000 controls r 2 =1.01,250 controlsr 2 = 0.80 Example: Tagger and LDSelect Using LD and Haplotypes to Pick tagSNPs Discovery genotype datapair-wise LDpick tagSNPs

40. LDSelect: Using LD to Pick tagSNPs LDSelect Uses SNP discovery data (not haplotypes) Finds all correlated SNP genotypes to minimize the total number Maintains genetic diversity of locus Carlson et al. AJHG (2004)

41. TagSNPs Are Population Specific European-descent (BLM) African-descent (BLM)

42. SNP Selection: tagSNP Data BLM

43. Side Note: Categorizing tagSNPs SNP context Nonrepetitive > repetitive Location of SNP Coding > noncoding Function Nonsynonymous > synonymous

44. Categorizing tagSNPs LPO

45. Haplotypes in Genetic Association Studies Two main approaches with haplotypes: HaplotypesPick tagSNPsGenotype samples Pick tagSNPs Infer haplotypesTest for association

46. Haplotypes in Genetic Association Studies Two main approaches with haplotypes: Haplotypes Pick tagSNPs Genotype samples Pick tagSNPs Infer haplotypesTest for association Recombination Natural selection Population history Population demography Haplotype block definition

47. Haplotype “Blocks” Strong LD Few Haplotypes Represent most chromosomes Daly et al 2001 Daly et al Nat. Genet. (2001)

48. Block Definitions Daly et al 2001 D ´ [Gabriel et al Science (2002)] Daly et al Nat. Genet. (2001)

49. Block Definitions AB ab Ab aB Four-gamete test: A B ab <4 haplotypes, D´=1block 4 haplotypes, D´<1boundary

50. Haplotype Blocks and tagSNPs Identifying blocks and tagSNPs: Manually Visual haplotype Algorithms HapMap and Haploview

51. Haplotype Blocks and tagSNPs LTA: 16 SNPs (MAF >10%) 6 “common” haplotypes tagSNPs

52. Haplotype Blocks and tagSNPs Identifying blocks and tagSNPs: Manually Visual Haplotype Algorithms HapMap and HaploView

53. HapMap Data and Haploview www.hapmap.org

56. HapMap Data and Haploview

59. http://www.broad.mit.edu/mpg/haploview/

60. Import HapMap Data into Haploview

68. Note: HapMap is not complete variation data

69. HapMap 5 tagSNPs Variation data, LD, and tagSNPs for ANAPC10 in European-Americans NIEHS SNPs 12 tagSNPs

70. tagSNPs and Genome Variation Server

72. Note: Tagger is essentially the same as LDSelect

74. Haplotypes, TagSNPs, and Caveats Haplotypes are inferred Block-like structure assumed for some software Different block definitions Block boundaries sensitive to marker density Genotype savings may not be great (recombination) tagSNPs based on LD more popular than htSNPs

75. Resources available for pair-wise LD and haplotypes Software for tagSNP selection available Be aware the limitations of the approach you choose Be aware that some SNP datasets may not represent all common variation of gene or gene region Be aware that a fraction of tagSNPs do not convert into a successful genotyping assay SNP Selection Summary