1Computational GenomicsComputational GenomicsPopulation Genetics:Population Genetics:SNPS Haplotype Inference SNPS Haplotype Inference Eric XingEric XingLecture 5, January 30, 2007Reading: hand-outs Genetic Polymorphisms2Type of polymorphismsz Insertion/deletion of a section of DNAz Minisatellites: repeated base patterns (several hundred base pairs)z Microsatellites: 2-4 nucleotides repeatedz Presence or absence of Alu segmentsz Single base mutation (SNP)z Restriction fragment length (RFLP)z Creating restriction sites via PCR primerz Direct sequencingFrequency of SNPs greater than that of any other type of polymorphismVariable Number of Tandem Repeats (VNTR) Polymorphism3Restriction Fragment Length Polymorphism (RFLP) Single Nucleotide Polymorphism (SNP)Polymorphism rate: number of letter changes between two different members of a speciesHumans: ~1/1,0004Exploiting Genetic Variationsz Population Evolution: the majority of human sequence variation is due to substitutions that have occurred oncein the history of mankind at individual base pairsz There can be big differences between populations!z Markers for pinpointing a disease: certain polymorphisms are in "Linkage Disequilibrium" with disease phenotypesz Association study: check for differences in SNP patterns between cases and controls z Forensic analysis: the polymorphisms provide individual and familiar signaturesa diploid individualCpCmchromosome “Binary” nt-substitutions at a single locus on a chromosome− each variant is called an "allele"GATCTTCGTACTGAGTGATCTTCGTACTGAGTGATTTTCGTACGGAATGATTTTCGTACTGAGTGATCTTCGTACTGAATGATTTTCGTACGGAATGATTTTCGTACGGAATGATCTTCGTACTGAATSingle Nucleotide Polymorphism (SNP)5Some Facts About SNPsz More than 5 million common SNPs each with frequency 10-50% account for the bulk of human DNA sequence differencez About 1 in every 600 base pairsz It is estimated that ~60,000 SNPs occur within exons; 85% of exons within 5 kb of nearest SNPGATCTTCGTACTGAGTGATCTTCGTACTGAGTGATTTTCGTACGGAATGATTTTCGTACTGAGTGATCTTCGTACTGAATGATTTTCGTACGGAATGATTTTCGTACGGAATGATCTTCGTACTGAATCTG 3/8TGA 3/8CTA 2/8Haplotype Consider J binary markers in a genomic region There are 2Jpossible haplotypes − but in fact, far fewer are seen in human population Good genetic marker for population, evolution and hereditary diseases …chromosomedisease XhealthyhealthyWhat is a haplotype?-- a more discriminative state of a chromosomal region62 136191517419629 172 121271467118 181410 10Genotypes Haplotypes131154921712761184102691716921214718110HaplotypeRe-constructionChromosome phase is knownChromosome phase is unknownHaplotype and Genotypez A collection of alleles derived from the same chromosomeLinkage Disequilibriumz LD reflects the relationship between alleles at different loci.z Alleles at locus A: frequencies p1,…, pmz Alleles at locus B: frequencies q1,…,qnz Haplotype frequency for AiBj:z equilibrium value: pi qjz Observed value: hijz Linkage disequilibrium: hij-pi qjz Linkage disequilibrium is an allelic association measure (difference between the actual haplotype frequency and the equilibrium value)z More precisely: gametic associationz Association studies.z If inheriting a certain allele at the disease locus increases the chance of getting the disease, and the disease and marker loci are in LD, then the frequencies of the marker alleles will differ between diseased and non-diseased individuals.7Use of Polymorphism in Gene Mappingz 1980s – RFLP marker mapsz 1990s – microsatellite marker mapsAdvantages of SNPs in geneticanalysis of complex traitsz Abundance: high frequency on the genomez Position: throughout the genome (level of influence of type of SNP, e.g. coding region, promoter site, on phenotypic expression?)z Ease of genotypingz Less mutable than other forms or polymorphismsz Allele frequency drift (different populations)z Haplotypic patterns8Haplotype analysesz Haplotype analysesz Linkage disequilibrium assessmentz Disease-gene discoveryz Genetic demographyz Chromosomal evolution studiesz Why Haplotypesz Haplotypes are more powerful discriminators between cases and controls in disease association studiesz Use of haplotypes in disease association studies reduces the number of tests to be carried out.z With haplotypes we can conduct evolutionary studies9CTATGACGATTA??????haplotype h≡(h1, h2)possible associations of alleles to chromosomeHeterozygousdiploid individualCTATGACpCmGenotype gpairs of alleles with association of alleles to chromosomes unknownATGCsequencingTC TG AAPhase ambiguity-- haplotype reconstruction for individualsInferring Haplotypesz Genotype: AT//AA//CGz Maternal genotype: TA//AA//CCz Paternal genotype: TT//AA//CGz Then the haplotype is AAC/TAG.z Genotype: AT//AA//CGz Maternal genotype: AT//AA//CGz Paternal genotype: AT//AA//CGz Cannot determine unique haplotypez Problem: determine Haplotypes without parental genotypes10Genotype representations0/0 Æ 01/1 Æ 10/1 Æ 2Genotypes of 14 individual21 2 222 0202 1 111 2211 0 000 0102 1 111 2221 2 222 0202 1 111 2211 0 000 0102 1 111 2221 2 222 0222 2 222 2121 1 222 0202 1 111 2222 2 222 2121 2 222 02|| | ||| ||Identifiability01 1 000 0011 0 000 0101 1 000 0000 1 111 1111 0 000 0111 0 000 0101 1 000 0000 1 111 1101 1 000 0011 0 000 0100 1 111 1101 1 000 0011 0 000 0111 0 000 0101 1 000 0000 1 111 1101 1 000 0011 0 000 0100 1 111 1111 0 000 0111 0 000 0101 1 000 0000 1 111 1101 1 000 0011 0 000 0100 1 111 1101 1 000 0011 0 000 0111 0 000 01|| | ||| || 01 1 000 00|| | ||| ||00 1 111 11|| | ||| ||1110701 1 111 0011 0 000 0101 1 111 0000 1 111 1111 0 000 0111 0 000 0101 1 111 0000 1 111 1101 1 111 0011 0 000 0100 1 111 1101 1 111 0011 0 000 0111 0 000 0101 1 111 0000 1 111 1101 1 111 0011 0 000 0100 1 111 1111 0 000 0111 1 000 0101 1 111 0000 1 111 1101 1 111 0011 0 000 0100 1 111 1101 1 111 0011 0 000 0111 0 000 01|| | ||| || 11 0 010 01|| | ||| ||11 1 000 01|| | ||| ||11 0 000 11|| | ||| ||01 1 111 00|| | ||| ||01 1 101 00|| | ||| ||01 0 111 00|| | ||| ||00 1 111 11|| | ||| ||00 1 111 01|| | ||| ||81118116101 1 101 0011 0 010 0101 1 111 0000 1 111 1111 0 000 0111 0 000 0101 1 111 0000 1 111 1101 0 111 0011 1 000 0100 1 111 1101 1 111 0011 0 000 0111 0 000 0101 1 111 0000 1 111 1101 1 111 0011 0 000 0100 1 111 1111 0 000 0111 1 000 0101 1 111 0000 1 111 1101 1 111 0011 0 000 1100 1 111 0101 1 111 0011 0 000 01??Parsimonious solutionIdentifiability11Three Problemsz Frequency estimation of all possible haplotypesz Haplotype reconstruction for