Docstoc

The Future of Behavioural Genetics in a Postgenomics World

Document Sample
The Future of Behavioural Genetics in a  Postgenomics World Powered By Docstoc
					           12th SGDP Summer School
                 20-24 June 2011

– Twin model-fitting: Introducing the new OpenMx
– Bioinformatics for Geneticists
– Introduction to R




        Thalia Eley          Fruhling Rijsdijk
         Director            Deputy Director
                  9am lectures
   Epigenetics
    – Jon Mill
   Gene X Environment in mental health
    – Avshalom Caspi
   GWAS: Contrasting findings from
    depression and Crohn’s disease
    – Catherine Lewis
   GWAS: A window into the genetic
    architecture of complex diseases
    – David Collier
                S+G+D+P
Robert Plomin, Deputy Director, SGDP Centre
Institute of Psychiatry, King’s College London

               Developmental




      Social                    Genetic
    D           S = Nurture (environment)
                G = Nature (genetics)
S        G      D = Development
                P = Behaviour

    • Nature and nurture
         -- Importance of both

    • Beyond nature vs nurture
         -- S, G, D, P, and S+G+D+P examples
     Why quantitative genetics?

• QG can estimate the cumulative effect of
  genetic influence regardless of the number
  of genes involved or the magnitude or
  complexity of their effects.
• QG can go beyond estimates of heritability
• Unlike MG, QG is as much a study of
  environment as genetics

              Haworth & Plomin (2010).
              Quantitative genetics in the era of molecular genetics:
              Learning abilities and disabilities as an example.
              J of Am Acad of Child & Adolescent Psychiatry
              (Special issue on genetics)
    D           S = Nurture (environment)
                G = Nature (genetics)
S        G      D = Development
                P = Behaviour

    • Nature and nurture
         -- Importance of both

    • Beyond nature vs nurture
         -- S, G, D, P examples
      5th edition of
 Behavioral Genetics
  (Plomin et al 2008)*

= genetics of behaviour
     quantitative genetics
     and molecular genetics



*available electronically
(robert.plomin@kcl.ac.uk)
 Quantitative genetics: twin method


  Identical twins    Fraternal twins
(monozygotic, MZ)    (dizygotic, DZ)
   Twin studies: Assess rather than
      assume genetic influence

How heritable are common medical
 disorders?

  – Parkinson’s disease
  – Breast cancer
  – Hypertension
  – Rheumatoid arthritis
  – Epilepsy
  – Peptic ulcers
   Twin studies: Assess rather than
      assume genetic influence

How heritable are common medical
 disorders?
                           Heritability?
  – Parkinson’s disease      ++++
  – Breast cancer            ++++
  – Hypertension              +++
  – Rheumatoid arthritis       ++
  – Epilepsy                    +
  – Peptic ulcers               -
Common medical disorders:
Concordances for MZ and DZ twins




                         Plomin et al. (1994) Science
Common behavioural disorders:
  Concordances for MZ and DZ twins




                         Plomin et al. (1994) Science
Normal dimensions of behaviour:
   Correlations for MZ and DZ twins




                           Plomin et al. (1994) Science
Perceptions of nature/nurture: % who think that
     nature is at least as important as nurture




                                     Walker & Plomin, 2005
            Nature and nurture

   Genetic influence is significant and substantial
    for nearly all behavioural traits and disorders
    – ‘heritable’ means DNA sequence variation

   Genetic research provides the best evidence
    for the importance of the environment
    – ‘environmental’ means everything other than DNA
      sequence variation
    – the way the environment works is surprising

   Need to go beyond heritability estimates
    – examples of S, G, D, and P
    D           S = Nurture (environment)
                G = Nature (genetics)
S        G      D = Development
                P = Behaviour

    • Nature and nurture
         -- Importance of both

    • Beyond nature vs nurture
         -- S, G, D, P examples from QG
       D

                                  S
S          G


   QG research provides best available
    evidence for the importance of
    environment

    – But the way the environment is surprising
       D

                                                   S
S              G

   The reason for QG designs is to separate
    nature from nurture in family resemblance
    – Genetics accounts for nearly all family
      resemblance
    – Nonetheless the environment is important
    – The salient aspects of the environment are those
      that do not contribute to family resemblance
          Nonshared environment
            – effects not events (e.g., divorce)
      Adoption design
    “Genetic-plus-environmental
             relatives”

                                  Shared
             Adoption         Environment (C)



“Genetic”               “Environmental”
 relatives                 relatives
Family and adoption results for weight:
      parents and their children




                         Direct estimate of
                         shared environment (C)
Family and adoption results for weight:
               siblings




                       Direct estimate of
                       shared environment (C)
Adult twin results for weight

            Direct estimate of non-
            shared environment (E)

                        ½ heritability




                      Shared environment
    Weight: nature and nurture
Shared environment    Nonshared environment




            Genetic
Environmental theories of weight

   Assume environmental influences shared by
    children growing up in the same family
    – nutrition and lifestyle


   But growing up in the same family does not
    make children similar in weight
     D

                                  S
S         G

   Nonshared environment accounts for
    most environmental influence
    – Psychopathology
    – Personality
    – Cognitive abilities after adolescence
    D


S        G
                                           S

       Why are two children growing up in the same
        family so different?
        – Study more than one child per family
        – MZ twins are especially useful


       Slow progress in identifying E
        – Biomarkers of non-shared environment?


                Plomin (2011). Why are children in the same family so
                different? Nonshared environment three decades later.
                International Journal of Epidemiology
           Nature and Nurture
   Genetics is important for nearly all traits

   Genetic research provides the best
    evidence for the importance of
    environment

   Nonshared environment accounts for
    most environmental influence
       D

                                  S
S          G


   QG research provides best available
    evidence for the importance of
    environment

    – But the way the environment is surprising:
      nonshareed environment
      D

                             D
S          G


   Developmental changes and continuity
    in genetic influences
General cognitive ability (‘g’, intelligence)
   is what diverse cognitive abilities
            have in common


        Verbal                 Spatial
                    ‘g’

                                  ~ 40% of
                                  the variance
                  Memory
                                      Twin studies of ‘g’
                                          (>10,000 pairs)


                       0.9
                       0.8                                               h2 ~ 50%
Correlations for ‘g’




                       0.7
                       0.6
                       0.5
                       0.4
                       0.3   c2 ~
                       0.2   35%
                       0.1
                        0
                                    MZ reared together   DZ reared together
                             MZ reared together vs
                              MZ adopted apart

                       0.9                    c2 ~ 10%
                       0.8                                           h2 ~ 75%
Correlations for ‘g’




                       0.7
                       0.6
                       0.5
                       0.4
                       0.3
                       0.2
                       0.1
                        0
                             MZ reared together   MZ adopted apart
                                    Adoption design
                                    (>1000 families)

                        0.5
                       0.45
Correlations for ‘g’




                        0.4
                       0.35                h2 ~ 50%
                        0.3
                       0.25                                          Parent-offspring
                        0.2                                          Sibling
                       0.15
                        0.1                                         c2 ~ 20%
                       0.05
                          0
                              Together   Adopted-apart   Adoptive
Meta-analyses of QG results for ‘g’

Nonshared
environment
    25%
                                 Genetic
                                  50%


     Shared
   environment
       25%

                 Chipuer, Rovine & Plomin (1990)
                  Genetic influence on ‘g’
              increases during development

               Error                                     Error
               10%                                        5%

                               Genetic
                                40%
                                                                            Genetic
Nonshared                                                                    60%
Environment                               Nonshared
    25%                                   Environment
                                              35%


          Shared Environment
                 25%

              Childhood                           After adolescence

                                  Plomin (1986) Development, Genetics & Psychology
                Heritability of ‘g’ increases
                   during the life span
               100
               90
               80
               70
               60
heritability




               50
               40
               30
               20
               10
                0
                     infant   child   adolescent   adult
Mega-analysis of 11,000 pairs of twins
        from four countries
 Heritability




Shared environment




   Haworth et al (2009) Molecular Psychiatry.
 Heritability




Shared environment




   Haworth et al (2009) Molecular Psychiatry.
Heritability increases across the life span
      for ‘g’ (and other traits too)
                 100
                 90
                 80
                           Why? GE correlation?
                 70
                 60
  heritability




                 50
                 40
                 30
                 20
                 10
                  0

                  infant     child   adolescent   adult
       D

                                    D
S           G


   Developmental changes and continuity
    in genetic influences
    – Heritability of ‘g’ increases throughout the
      lifespan
       D

                              G
S          G


   Genetics of co-morbidity and
    heterogeneity
Univariate genetic analysis


            hX        G


  Trait X

                 ex

                      E
Univariate genetic analysis


           hX        G


 Reading

                ex

                     E
Univariate genetic analysis

   G        hY




                 Trait Y
            eY


   E
Univariate genetic analysis

   G        hY




                 Maths
            eY


   E
   Multivariate genetic analysis


     hX   G            G    hY
                rG

Reading                      Maths


     eX         rE          eY
          E            E
  Multivariate genetic analysis

          ‘Bivariate heritability’


   hX     G                 G        hY
                  rG

Reading                              Maths


   eX              rE                eY
          E                 E
  Multivariate genetic analysis
          Genetic correlation


   hX     G               G     hY
                 rG

Reading                         Maths


   eX            rE             eY
          E               E
          Phenotypic overlap




Reading                                    Maths




           Phenotypic correlation ~ 0.60
          Phenotypic overlap
            = 65% genetic


Reading                 65%                 Maths
                       genetic




65% of phenotypic covariance is mediated genetically
              (‘bivariate heritability’)
             Genetic overlap



  Genetic                                   Genetic
                        80%               influences
influences
on reading                                 on maths


             Genetic correlation ~ 0.80

           Not just reading and maths:
Similar results for language, cognitive abilities, ‘g’
Davis, Haworth & Plomin (2009)
                Generalist genes

• Disabilities as well as abilities
    Generalist Genes Hypothesis

 Although there are specific genes,
  most of the genes that affect one
  ability/disability affect all.
 In other words, one set of genes
  affects diverse learning abilities and
  disabilities.
            Plomin & Kovas (2005).
            ‘Generalist genes’ and learning disabilities.
            Psychological Bulletin.
    Generalist Genes Hypothesis

   For other behavioural domains too
    – Psychopathology
    – Not for all (e.g., autism)
   General phenomenon?
    – Pleiotropy: Each gene affects many
      traits
    – Polygenicity: Each trait is affected by
      many genes
    Generalist Genes Hypothesis

 One set of genes affects diverse
  learning abilities and disabilities
 Implications for psychology1,
  neuroscience2, and molecular
  genetics3

1 Kovas & Plomin (2007) Current Directions in Psych Science
2 Kovas & Plomin (2006) Trends in Cognitive Science
3 Butcher et al (2006) Current Opinion in Neurobiology
       D

                              G
S          G


   Genetics of co-morbidity and
    heterogeneity
    – Generalist genes
      D

                             P
S          G


   The relationship between the abnormal
    and normal (disorders vs dimensions)
Two worlds of genetics: QG and MG
       Quantitative    Molecular
        Genetics       Genetics

1900
       Biometricians   Mendelians

1950


1975
1985

2000
Two worlds of genetics: QG and MG



1860

1870                      Gregor Mendel

         Francis Galton

1900
          Quantitative      Molecular
           Genetics         Genetics
            Gregor Mendel
                (1866)
                                     Edible pea (Pisum sativum)
A single gene with two alleles can
   explain the inheritance of a
qualitative (dichotomous) disorder
                          Francis Galton (1877)

                          Quantitative traits can be
                      inherited from parent to offspring


    Sweet pea
(Lathyrus odoratus)
Two worlds of genetics: QG and MG
       Quantitative          Molecular
        Genetics             Genetics

1900
       Biometricians         Mendelians

1950   Multiple genes        Single genes
       Quantitative traits   Qualitative traits
       Phenotype-centred     Gene-centred
       Naturally occurring   Experimentally induced
1975
1985

2000
                               Fisher’s resolution




 R.A. Fisher (1918)

      Mendel’s laws of
inheritance for qualitative
       traits apply to
 quantitative traits if they
 are influenced by several
           genes.
          What about disorders?

   Fisher: Mendelian inheritance of qualitative
    traits can account for the inheritance of a
    quantitative trait if many genes are involved

   What about converse: qualitative traits
    (disorders) influenced by many genes?
Two worlds of genetics: QG and MG
       Quantitative      Molecular
        Genetics         Genetics

1900



1950


1975
1985

2000
       Genome-wide association (GWA)
    Common disorders are influenced by
       many genes of small effect:
     Genome-wide association studies
   GWA has identified hundreds of associations with
    complex disorders 1
   But effect sizes are very small 2
    – Risk < 2 for complex disorders
   ‘Missing heritability’   3

   How to proceed? 4
                  1 Hindorff(2009) PNAS
                  2 McCarthy (2008) Nature Reviews Genetics
                  3 Maher (2008) Science
                  4 Manolio (2009) Nature
             Quantitative trait locus (QTL)
               hypothesis: a syllogism



1. Multiple genes  quantitative traits
2. Case-control GWAS of common disorders
    indicate multiple genes
3. Common disorders are quantitative traits
Cognitive disability: 282 monogenic effects
           (Inlow & Restifo, 2004, Genetics)
Fragile X                  Gene-1
   •.001 (-2 SD IQ)                                 < 1%
Rett                         …




                                                        “missing heritability”
   • .0005 (-3 SD IQ)      Gene-282
Neurofibromatosis
   • .0003 (-1 SD IQ)
Duchenne musc. dystrophy                    ?
   • .0003 (-1 SD IQ)
Phenylketonuria (PKU)
   • .0001 (-3 SD IQ)
Lesch-Nyhan
   • .00005 (-3 SD IQ)                Heritability of
                                        cognitive
= rare and severe                       disability
Quantitative trait locus (QTL) hypothesis
What we call disorders are the quantitative extremes
 of the same genetic (and environmental) factors
          responsible for normal variation



    Cut-off for
     disorder




                   Normal variation
             QTL hypothesis:
     Case-control associations are QTLs



      Cut-off for
       disorder


    ‘Cases’                                 ‘Controls’
= 70% A1A1 (red)                        = 48% A1A1 (red)




         Normal variation in quantitative trait
        QTL hypothesis:
Case-control associations are QTLs




   Normal variation in quantitative trait
     Implications of QTL hypothesis:
          Polygenic risk scores
          (composites of many SNPs)



•   Problem
•   Disability
•   Disabling
•   Vulnerability




             Quantitative trait variation
Novel implication of QTL hypothesis:
         Positive genetics



•   Problem                            •   Positive
•   Disability                         •   Ability
•   Disabling                          •   Enabling
•   Vulnerability                      •   Resilience




              Quantitative trait variation
       The QTL hypothesis:
Common disorders are quantitative traits




           Quantitative trait variation
  Taking the QTL hypothesis seriously

From:
   GWA of                          Quantitative
common disorders        QTLs          traits

 To:
    GWA of                          Common
 quantitative traits    QTLs        disorders

Bio-QTs: Transcriptomics, epigenomics, metabolomics
      D

                                  P
S          G


   The relationship between the abnormal
    and normal (disorders vs dimensions)
    – Common disorders are quantitative traits
    D           S = Nurture (environment)
                G = Nature (genetics)
S        G      D = Development
                P = Behaviour

    • Nature and nurture
         -- Importance of both

    • Beyond nature vs nurture
         -- S, G, D, P examples

    • Interplay between nature and nurture
    D            S = Nurture (environment)
                 G = Nature (genetics)
S         G      D = Development
                 P = Behaviour

        • Nature and nurture
          -- Importance of both

        • Beyond nature vs nurture
          -- S, G, D, P examples

    • Interplay between nature and nurture
          -- S + G + D + P
      D

                       S+G+D+P
S          G


   Gene-environment interaction and
    correlation
     Going beyond nature versus nurture:
    Interplay between nature and nurture


   Gene-environment interaction
     – Genetic sensitivity to environments
            Candidate genes and GxE
Role of genotype in the cycle of violence in maltreated children
Author(s): Caspi A, McClay J, Moffitt TE, et al.
Source: SCIENCE Volume: 297 Issue: 5582 Pages: 851-
854 Published: AUG 2 2002
Times Cited: 1,168

Influence of life stress on depression: Moderation by a polymorphism
in the 5-HTT gene
Caspi A, Sugden K, Moffitt TE, et al.
SCIENCE Volume: 301 Issue: 5631 Pages: 386-389
Published: JUL 18 2003
Times Cited: 2,200
                                         GxE for depression symptoms
                             12.50
Self reports of depression
    symptoms, age 26



                             10.00


                              7.50


                              5.00                                 5-HTTLPR
                                                                    SS, n = 146
                                                                    SL, n = 435
                              2.50
                                                                    LL, n = 264


                              0.00
                                     0        1      2       3       4+

                                          Number of life events, ages 21-26


                                                                 Caspi et al (2003) Science
     Going beyond nature versus nurture:
    Interplay between nature and nurture


   Gene-environment interaction
     – Genetic sensitivity to environments


   Gene-environment correlation
     – Genetic influence on exposure to environments
          Three types of GE correlation
                                         Anything
                              Active
Passive




             G
                      Child              Environment
             E
                              Reactive

Relatives                                 Anyone

                       Plomin, DeFries & Loehlin (1977)
                       Psychological Bulletin
Traditional model of E measures




                G
                        Behavioral
Environmental           Phenotype
  Measure       E
           GE correlation model

            G
                       Environmental
                         Measure
            E


Heritability of the environment?
   Measures of environment often involve
    behavior
   Analyze E measures as dependent
    measures in quantitative genetic designs
Antecedents of genetic influence on E measures

     Traditional view
                             G
                                    Behavioral
             Environmental          Phenotype
                             E
               Measure


     GE correlation


 G                           G
              Behavioral           Environmental
              Phenotype              Measure
 E                           E
Sequelae of genetic influence on E measures

    Traditional view
                            G
                                  Behavioral
            Environmental         Phenotype
                            E
              Measure


    GE correlation


G                           G
            Environmental         Behavioral
              Measure             Phenotype
E                           E
General twin analysis
 Use Plomin et al 96




       Pike … Plomin (1994) Developmental Psychology
Genetic factors mediate associations
 between environmental measures
     and behavioral outcomes

   Family environment and psychopathology
   Family environment and personality
   Family environment and IQ
   Social support and depression
   SES and IQ
   SES and academic achievement
   Peer deviance and conduct disorder
      DNA studies of GE correlation:
               Predictions

   DNA variants are associated with measures
    of the environment

   DNA variants mediate associations between
    environmental measures and behavioural
    measures
       DNA studies of GE correlation:
               Prediction 1


DNA variants are associated with measures of
              the environment

 DNA


  G
           Environmental
             Measure
  E
      DNA studies of GE correlation:
              Prediction 2
      DNA variants mediate associations
    between environmental measures and
           behavioural measures
DNA


G                          G
           Environmental            Behavioral
             Measure                Phenotype
E                          E
      GE correlation: Conclusion

   Genetics contributes substantially to
    measures of the environment and to
    the association between measures of
    the environment and behaviour
   GE correlation: Implication

From passive model of environment
to active model of experience

  – People select, modify, construct and
    re-construct their experiences in part
    on the basis of their genetic propensities
      D

                       S+G+D+P
S          G


   Gene-environment correlation
    D           S = Nurture (environment)
                G = Nature (genetics)
S        G      D = Development
                P = Behaviour

    • Nature and nurture
         -- Importance of both

    • Beyond nature vs nurture
        D       S = Nurture (environment)
                G = Nature (genetics)
    S       G   D = Development
                P = Behaviour

 S (E but not C)
 G (Generalist genes)
 D (Heritability increases)
P   (Common disorders are QTs)
 S+G+D+P (Gene-environment correlation)

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:4
posted:3/18/2012
language:English
pages:102