An quick overview ofhuman genetic linkageanalysisStat 246, Lecture 2, Part APurpose of humanlinkage analysisTo obtain a crude chromosomallocation of the gene or genesassociated with a phenotype ofinterest, e.g. a genetic disease oran important quantitative trait.Examples: cystic fibrosis (found),diabetes, multiple sclerosis, andblood pressureLinkage StrategiesTraditional (from the 1980s or earlier)– Linkage analysis on pedigrees– Allele-sharing methods: candidategenes, genome screen– Association studies: candidategenes– Animal models: identifyingcandidate genesNewer (from the 1990s)– Focus on special populations(Finland, Hutterites)– Haplotype-sharing (many variants)– Congenic/consomic lines in mice(new for complex traits)Linkage analysisAllele-sharingmethodsAssociation StudiesAnimal ModelsLinkage Strategies IIOn the horizon (here)– Single-nucleotide polymorphism(SNPs)– Functional analyses: findingcandidate genesNeeded (starting to happen)– New multilocus analysis techniques,especially– Ways of dealing with largepedigrees– Better phenotypes: ones closer togene products– Large collaborationsHorses for courses• Each of these strategies hasits domain of applicability• Each of them has a differenttheoretical basis and methodof analysis• Which is appropriate formapping genes for a diseaseof interest depends on anumber of matters, mostimportantly the disease, andthe population from which thesample comes.The disease mattersDefinition (phenotype),prevalence, features such asage of onsetGenetics: nature of genes(resistance, susceptibility),number of genes, nature oftheir contributions (additive,interacting), size of effectEnvironment, other relevantvariables (e.g. sex)Genotype-by-environmentinteractionsThe population mattersHistory: pattern of growth,immigrationComposition: homogeneous ormelting pot, or in betweenMating patterns: family sizes,mate choice (level ofconsanguinity)Frequencies of disease-relatedalleles, and of marker allelesAges of disease-related allelesComplex traitsDefinition vague, but usually thought of ashaving multiple, possibly interacting loci,with unknown penetrances; andphenocopies. The terms polygenic andoligogenic are also used, but these do havemore specific meanings.There is some evidence that using a range ofmade-up models can help map genes forcomplex traits, but no-one really knows.Affected only methods are widely used, withvariance component methods becomingpopular. The jury is still out on which, if anywill succeed.Few success stories so far.Important: heart disease, cancer susceptibility,diabetes, …are all “complex” traits.Design of genemapping studiesHow good are your data implying agenetic component to your trait? Canyou estimate the size of the geneticcomponent?Have you got, or will you eventuallyhave enough of the right sort of data tohave a good chance of getting adefinitive result?Power studiesSimulations.GenotypingChoice of markers: highlypolymorphic preferred.Heterozygosity and PICvalue are the measurescommonly used.Reliability of markersimportant tooGood quality data critical:errors can play a surprisinglylarge role.Preparing genotypedata for analysisData cleaning is thebig issue here.Need much ancillarydata…how good is it?AnalysisA very large range ofmethods/programs areavailable.Effort to understand theirtheory will pay off inleading to the right choiceof analysis tools.Trying everything is notrecommended, but notuncommon.Many opportunities forinnovation.Interpretation ofresults of analysisAn important issue here is whetheryou have established linkage. Thestandards seem to be gettingincreasingly stringent.What p-value or LOD should youuse?Dealing with multiple testing,especially in the context of genomescans and the use of multiplemodels and multiple phenotypes, isone of the big issues.Replication of resultsThis has recently become a bigissue with complex diseases,especially in psychiatry.Nature Genetics suggested in May1998 that they will requirereplication before publishing resultsmapping complex traits.Simulations by Suarez et al (1994)show that sample sizes necessaryfor replication may be substantiallygreater than that needed for firstdetection.Topics not mentioned Sex-linked traits, sex-specificrecombination fractions, liability classes,mutations, genetic heterogeneity,exclusion mapping, homozygositymapping, interference, variancecomponent methods, twin studies, andmuch more. Some of these topics plus the ones arecovered in two books: Handbook of Human Genetic Linkageby J.D. Terwilliger & J. Ott (1994)Johns Hopkins University Press Analysis of Human Genetic Linkageby J. Ott, 3rd Edition (1999),Johns Hopkins University
View Full Document
Unlocking...