12
Genetics and Analysis of Quantitative Traits Michael Lynch Bruce Walsh

Genetics and Analysis of Quantitative Traits

Embed Size (px)

DESCRIPTION

Genetics and Analysis of Quantitative Traits

Citation preview

  • Genetics andAnalysis ofQuantitative TraitsMichael LynchBruce Walsh

  • Contents

    CONTENTS i

    FREFACE xiii

    I. FOUNDATIONS OF QUANTITATIVE GENETICS i1. AN OVERVIEW OF QUANTITATIVE GENETICS 3

    The Adaptationist Approach to Phenotypic Evolution 3Quantitative Genetics and Phenotypic Evolution 4Histrica! Background 7The Major Goals of Quantitative Genetics 13

    The nature of quantitative-trait variation 13The consequences of inbreeding and outcrossing 14The constraints on the evolutionary process 15The estimation of breeding vales 15The developrnent of predictive models for evolutionary change 16

    Mathematics in Biology 16

    2. PROPERTIES OF DISTRIBUTIONS 19Parameters of Univariate Distributions 19The Normal Distribution 26

    The truncated normal distribution 29Confdence Intervals 32

    3. COVARIANCE, REGRESSION, AND CORRELATION 35Jointly Distributed Random Variables 35

    Expectations of jointly distributed variables 36Covariance 36

    Useful identities for variances and covariances 38Regression 39

    Derivation of the least-squares linear regression 39Properties of least-squares regressions 41

    Correlation 43A Taste of Quantitative-Genetic Theory 45

    Directional selection differentials and the Robertson-Price identity 45The correlation between genotypic and phenotypic vales 47Regression of offspring phenotype on parental phenotype 48

  • ii CONTENTS

    4. PROPERTIES OF SINGLE LOCI 51Alele and Genotype Frequencies 52The Transmission of Genetic Information 54

    The Hardy-Weinberg principie 54Sex-linked loci 56Polyploidy 57Age structure 60Testing for Hardy-Weinberg proportions 60

    Characterizing the Influence of a Locus on the Phenotype 61The Basis of Dominance 63Fisher's Decomposition of the Genotypic Valu 65Partitioning the Genetic Va nance 69Additive Effects, Average Excesses, and Breeding Vales 71Extensions to Mltiple Aleles and Nonrandom Mating 74

    Average excess 74Additive effects 75Additive genetic variance 76

    5. SOURCES OF GENETIC VARIATION FORMULTILOCUS TRAITS 81

    Epistasis 82A General Least-Squares Model for Genetic Effects 85

    Extensin to haploids and polyploids 92Linkage 94

    Estimation of gametic phase disequilibrium 97Effect of Disequilibrium on the Genetic Variance 100

    The evidence 103

    6. COMPONENTS OF ENVIRONMENTAL VARIATION 107Extensin of the Linear Model to Phenotypes 108Special Environmental Effects 111

    Within-individual variation 112Developmental horneostasis and homozygosity 116Repeatability 121

    General Environmental Effects of Maternal Influence 123Genotype x Environment Interaction 127

    7. RESEMBLANCE BETWEEN RELATIVES 131Measures of Relatedness 132

    Coefficients of identity 133Coeffcients of coancestry and inbreeding 135The coefficient of fraternity 140

    The Genetic Covariance Between Relatives 141The Effects of Linkage and Gametic Phase Disequilibrium 146

  • CONTENTS iii

    Linkage 146Gametic phase disequilibrium 150

    Assortative Mating 153Polyploidy 161Environmental Sources of Covariance Between Relativas 162The Heritability Concept 170

    Evolvability 175

    8. INTRODUCTION OF MATRIX ALGEBRAAND LINEAR MODELS 177

    Mltiple Regression 177An application to multivariate selection 180

    Elementary Matrix Algebra 182Basic notation 182Partitioned matrices 183Addition and subtraction 183Multiplicador! 184Transposition 186Inverses and solutions to systems of equations 187Determinante and minors 189Computing inverses 190

    Expectations of Random Vectors and Matrices 192Covariance Matrices of Transformed Vectors 193The Multivariate Normal Distribution 194

    Properties of the MVN 195Overview of Linear Models 198

    Ordinary least squares 200Generalized least squares 202

    9. ANALYSIS OF LINE CROSSES 205Expectations for Line-cross Means 206Estimacin of Composite Effects 213

    Hypothesis testing 215Line crosses in Nicotiana rustica 219Additional data 221

    The Genetic Interpretation of Heterosis and Outbreeding Depression 222Variance of Line-cross Derivatives 226Biometrical Approaches to the Estimation of Gene Number 231

    The Castle-Wright estimator 233Effect of the leading factor 238Extensions to haploids 241

    Other Biometrical Approaches to Gene Number Estimation 244The inbred-backcross technique 244Genotype assay 246

  • iv CONTENTS

    10. INBREEDING DEPRESSION 251The Genetic Basis of Inbreeding Depression 252

    A more general model 256Methodologjcal Considerations 259

    Single-generation analysis 260Multigenerational analyses 262Ritland's method 266Epistasis and inbreeding depression 267Variance in inbreeding depression 268

    The Evidente 269Purging Inbreeding Depression 274Nurnber of Lethal Equivalents 276

    Rcsults from vertebrares 278Results from Dmsoplia 279Results from plants 281

    Partial Recessives vs. Overdominance 283The (A+B)/A ratio 283Estimating the average dcgree of dominance 284Inferences from molecular markers 287

    11. MATTERS OF SCALE 293Transformations to Achieve Normality 293

    Log-normal distributions and the log transform 294Tests for normality 295

    Stabilizing the Variance 300Kleckowski's transformation 300General variance stabilizing-transformations 301The Roginskii-Yablokov effect 302The Kluge-Kerfoot phenomenon 305

    Allometry: the Scaling Implications of Body Size 305Removing Interaction Effects 307Developmental Maps, Canalization, and Genetic Assimilation 309

    Estimating developmental maps 310Selection and canalization 314Genetic assimilation 316

    11. QUANTITATIVE TRAIT LOCI 31912. FOLYGENES AND POLYGENIC MUTATION 321

    The Gene tic Basis of Quantitative-Genetic Variation 322Major genes and isoalleles 322The molecular nature of QTL variation 323

    The Mutational Rate of Production of Quantitative Variation 328

  • CONTENTS v

    Estirnation from divergente experiments 330Bristle numbers in Dwsophila 333Additional data 335

    The Deleterious Effects of New Mutations 340The Bateman-Mukai technique 341Results from flies, plants, and bacteria 343Analysis of natural populations 348The persistente of new mutations 351

    13. DETECTING MAJOR GENES 353Elementar}' Tests 354

    Departures from normality 354Tests based on sibship variances 355Major-gene ndices (MGI) 357Konpararnetric line-cross tests 358

    Mixture Models 359The distribution under a mixture model 360Parameter estimation 360Hypothesis testing 361

    Complex Segregation Analysis 364Likelihood functions assuming a single major gene 366Common-family effects 370Polygenic background 371Other extensions 373Ascertainment bias 374Estimating individual genotypes 374

    Analysis of Discrete Characters 375Single-locus penetrante model 376Major gene plus a polygenic background 377

    14. PRINCIPLES OF MARKER-BASED ANALYSIS 379Classical Approaches 379

    Chromosomal assays 380Thoday's method 381Genetics o Drosophila bristle number 385Genetics of Dwsophila speciation 387

    Molecular Markers 390Genetic Maps 393

    Map distance vs. recombination frequencies 394How many markers are needed? 397

    Marker-trait Associations 398Selective genotyping and progeny testing 401Recornbinant inbred lines (RILs) 401Bulked segregant analysis 402

  • vi CONTENTS

    QTL mapping by marker changes in populations under selection 404Marker-based Analysis Using Nearly Isogenic Lines (NILs) 405

    Marker-based introgressions 407Fine Mapping of Major Genes Using Population-level Disequilibrium 413

    LD mapping in expanding populations 414Candidate Loci 418

    The transmission/disequilibrium test 419Estimating effects of candidate loci 422Templeton and Sing's method: Using the histrica!

    information in haplotypes 424Cloning QTLs 425

    Transposon tagging 426Positional cloning and comparative mapping 426

    15. MAPPING AND CHARACTERIZING QTLS:INBRED LINE CROSSES 431

    Foundations of Line-Cross Mapping 431Experimental designs 432Conditional probabilities of QTL genotypes 433Expected marker-class means 437Marker variance and higher-order moments 439Overall significance leve! with mltiple tests 441

    QTL Detection and Estimation Using Linear Models 442QTL Detection and Estimation via Mximum Likelihood 445

    Likelihood maps 446Precisin of ML estimates of QTL position 448ML interval mapping 450Approximating ML interval mapping by Haley-Knott regressions 453

    Dealing with Mltiple QTLs 457Marker-difference regression 459Interval rnapping with marker cofactors 463Detecting mltiple linked QTLs using standard marker-trait regressions .. 467

    Sample Size Required for QTL Detection 469Power under selective genotyping 474Power and repeatabilty of mapping experiments 474

    Selected Applications 477The nature of transgressive segregation 477QTLs involved in reproductive isolation in Minndus 478QTLs involved in protein regulation 478QTLs in the Illinois long-term selection maize lines 479QTLs involved in the differences between maize and teosinte 481QTLs for age-specific growth in mice 484Summary o QTL mapping experiments 484

  • CONTENTS vii

    16. MAPPING AND CHARACTERIZING QTLS:OUTBRED POPULATIONS 491

    Measures of Informativeness 492Sib Analysis: Linear Models 495

    A single half-sib farnly 496Several half-sib families 498

    Power of Nested ANOVA Designs 501A single full-sib family 502Several full-sib families 504

    Sib Analysis: Mximum Likelihood 505Constructing likelihood functions 507

    Mximum Likelihood over General Pedigrees: Variance Componente 510Estimating QTL position 512

    The Haseman-Elston Regression 513Derivation of the Haseman-Elston regression 513Estimating the number of marker genes ibd 516Power and improvements 517Interval mapping by a modified Haseman-Elston regression 518

    Mapping Dichotomous Characters 521Recurrent and relative risks of pairs of relatives 523Affected sib-pair tests 525Power of ASP tests and related issues 527Genomic scanning 529Exclusin mapping and information content mapping 530Affected pedigree member tests 532

    III. ESTIMATION PROCEDURES 53517. PARENT-OFFSPRING REGRESSION 537

    Estimation Procedures 538Balanced data 538Unequal family sizes 539Standardizaron of data from the different sexes 542

    Precisin of Estimates 542Optimum Experimental Designs 543

    Assortative mating 547Estimation of Heritability in Natural Populations 548Linearity of the Parent-Offspring Regression 550

    18. SIB ANALYSIS 553Half-sib Analysis 554

    One-way analysis of variance 556Hypothesis testing 560

  • viii CONTENTS

    Sampling variance and standard errors 561Confidence nter vals 562Negativo estimares of heritability 563Optimal experimental design 564Unbalanced data 566Resampling procedures 569

    Full-sib Analvsis 570Nested analvsis of variance 573Hypothesis testing 574Sampling error 576Optimal design 577

    19. TWINS AND CLONES 581The Classical Approach 582

    Heritability estimation 584The Monozygotic-Twin Half-sib Method 587Clonal Analysis 592

    20. CROSS-CLASSIFIED DESIGNS 597North Carolina Design TI 598

    The average degree of dominance 603The Cockerham-Weir model 605

    Diallels 610Pooled reciprocis, no self crosses 611Reciprocis, no self crosses 614Complete diallels 618Partial diallels 618

    Hayman-Jinks analysis 619North Carolina Design III and the Triple Test Cross 625Some Closing Statistical Considerations 627

    21. CORRELATIONS BETWEEN CHARACTERS 629Theoretical Composition of the Genetic Covariance 630Estimation of the Genetic Covariance 632

    Pairwise comparison of relatives 632Nested analvsis of variance and covariance 633Regression of family means 636

    Components of Phenotypic Correlation 637Phenotypic correlations as surrogate estmales of genetic correlations . . . 639

    Statistical Issuos 639Hypothesis tests 641Standard errors 642Bias due to selection 644

    Applications 648

  • CONTENTS ix

    Genetic basis of population differentiation 648The homogeneity of genetic covariance matrices among species 650Evolutonary allometry 653Evolution of life-history characters 655

    22. GENOTYPE x ENVIRONMENT INTERACTION 657Genetic Correlation Across Two Environments 660

    Estimation procedures 663Two-way Analysis of Varance 666

    Relationship to Falconer's correlation across environments 671Further Characterization of Interaction Effects 672

    Joint-regression analysis 672Testing for Cross-over Interaction 678Concepts of Stability and Plasticity 680

    Additional issues 682The Quantitative Genetics of Genotype x Environrnent Interaction 683

    23. MATERNAL EFFECTS 687Components of Variance and Covariance 689

    Cytoplasmic transmission 693Postpollination reproductive traits in plants 695

    Cross-fostering experiments 696Body weight in mice 700

    Eisen's Approach 703Bondari's experiment 703

    Falconer's Approach 706Extensin to Other Types of Relatives 711

    24. SEX LINKAGE AND SEXUAL DIMORPHISM 715Sex-Hnked Loci and Dosage Compensation 715Sex-modified Expression of an Autosomal Locus 718

    Gametic imprinting 718Extensin to Mltiple Loci and the Covariance Between Relatives 719Variation for Sexual Dimorphism 724

    25. THRESHOLD CHARACTERS 727Heritability on the Underlying Scale 730Mltiple Thresholds 736Genetic Correlation Among Threshold Traits 739Heritability on the Observed Scaie 741

    26. ESTIMATION OF BREEDING VALES 745The General Mixed Model 746Estimating Fixed Factors and Predicting Random Effects 748

  • x CONTENTS

    Estimability o fxed factors 753Standard errors 754

    Models for he Estimation of Breeding Vales 755The animal model 755The gametic model 758The reduced animal model 759

    Simple Rules for Computing A and A"1 762Allowing for mutation when computing A 766

    Joint Estimation of Several Vectors of Random Effects 767BLUP estmales of dominance vales 767Repeated records 769Maternal effects 773Mltiple traite 774

    27. VARIANCE-COMPONENT ESTIMATION WITHCOMPLEX PEDIGREES 779

    ME versus REME Estimates of Variance Components 780A simple exaniple of ML versus REML 781

    ME Estimates of Variance Components in the General Mixed Model 784Standard errors of ML estimates 788

    Restricted Mximum Likelihood 789Multivariate analysis 792ML/REML estimation in populations undcr selection 792

    Solving ML/REML Equations 793Derivan've-based methods 794EM methods 797Additional approaches 799

    A Molecular-marker Based Method for Inferring Variance Components . . . . 800

    IV. APPENDICES 805

    Al. EXPECTATIONS, VARIANCES, AND COVARIANCESOF COMPOUND VARIABLES 807

    The Delta Method 807Expectations of complex variables 808Variances of complex variables 810Covariances of complex variables 813

    Variance of Variances and Covariances 813Expectations and Variances of Products 817Expectations and Variances of Ratios 818

    Sampling variance of regression and correlations coefficients 818Sampling variance of a coefficient of variation 819

  • CONTENT5 xi

    A2. PATH ANALYSIS 823Univariate Analysis 823Bivariate Analysis 826Applications 826

    Phenotypic correlation between parents and offspring 827Correlations between characters 829Growth analysis 831

    A3. FURTHER TOPICS IN MATRIX ALGEBRA ANDLINEAR MODELS 835

    Generalized Inverses and Solutions to Singular Systems of Equations 835Generalized inverses 836Consistency and solutions to consistent systems 836Estmability of fixed factors 839

    The Square Root of a Matrix 841Derivation of the GLS Estimators 842Quadratic Forms and Sums of Squares 843

    Moments of quadratic forms 843The sarnple variance as a quadratic form 844Sums of squares expressed as a quadratic form 846

    Testing Hypotheses About Linear Models 848Equivalent Linear Models 849Derivativas of Vectors and Matrices 851

    A4. MXIMUM LIKELIHOOD ESTIMATION ANDLIKELIHOOD-RATIO TESTS 853

    Likelihood, Support, and Score Functions 853Large-sample properties of MLEs 854The Fisher information matrix 855

    Likelihood-ratio tests 857The G-test 859Likelihood-ratio tests for the general linear model 860

    Iterativa Methods for Solving ML Equations 861Newton-Raphson methods 861Expectation-maximization methods 863EM for mixture model likelihoods 863EM rnodifications for QTL mapping 865

    A5. COMPUTING THE POWER OF STATISTICAL TESTS 869Power of Normally Distributed Test Statistics 870

    One-sided tests 870Two-sided tests 872Applications: Parent-offspring regressions 874