Main Topics Notes
Lec 1 - Introduction - genetic sensitivity RA & depression, structural brain changes, immune system, serotonin &
depression → an interaction between processes, multidimensional intervention
application - allergy: early manifestation of immune dysregulation, classic psychosomatic disorder, complex
interactions, socially relevant; higher rate within families (35-75% genes), environ: too quick
incidence to be evolution alone, influence of stress, epigenetics processes: immune cell
differentiation & regulation; time window & dose influences, heterogeneous
- schizophrenia: heritability gap (twin studies overestimate heredity bc also share environ;
(20-35% genes)
Down's: 100% genes; MDD: 20-25% genes (a er correction for shared environ)
- clinical: change environ; Crisp-CAS9 changes DNA and alters epigenetics in many ways
How to Use an Article - pharmacogenomics = genetics to improve diagnosis and therapy
about Genetic Association - pedigree = diagram w/ heritable traits across 2+ generations of a family
- allele = variant of a gene; haplotype = alleles occurring together bc SNPs are close (so are
inherited together = genetic linkage); variant allele = the last frequent one (VS wild-type allele)
- cell → nucleus → chromosomes (22 pairs + 1 sex, winded by histones = proteins, to fit) → DNA
(bars are sugar-phosphate) → genes (a sequence of letters that encode for protein: adenine,
thymine, guanine, cytosine; each base pair is a rung; nucleotide = 1 base + phosphate groups; starts
w/ promotor, ends with terminator region)
- genome: DNA set of a human, 23 chromosomes egg + 23 chromosomes sperm
- phenotype: GxE, outward charac, personality (outside gene), genotype: genetic make-up
- SNPs = single-nucleotide polymorphism (single base pair change, vs common sequence);
synonymous SNP = SNP does not lead to change in amino acid
- VNTRs = variable number of tandem repeats
- epigenome: what is actually activated/ deactivated, some inprints survive through generations,
is changeable, depends on timing, changes constantly throughout the day!;
- epigenetics: without changes in DNA sequence, biochemical processes that switch genes
on/off, triggered by environ factors, signals come from the inside cell, nearby cells, and exterior;
modifies histones and accessibility to DNA transcription
- isoform = variant in amino acid sequence of a protein, the different proteins
Mechanisms of Genetics - gene to protein: DNA is transcribed into mRNA (messenger ribonucleic acid) → mRNA translated
into protein, in the ribosome (protein synthesis machinery, outside nucleus); first it's unwinded
genetic variation - >99% of DNA is similar across humans, <1% differs across individuals
- mutations = rare permanent change in base (deletion/ duplication/ DNA not repaired), <1%
- polymorphisms = 2+ variants of a gene (allele) occurring w/ at least 1% of the less common
variant; transmitted through generations, no major defect in bio functions: 1) presence/ deletion
of a stretch of DNA (involves DNA duplication = copy number variation = CNV); 2) repeating
patterns of DNA; 3) single-base pair change (SNPs), may be from segment that codes for protein,
or just influence cell function, the most common
- e.g. APOE gene: 3 alleles = e2, e3 (most common in white population), e4; additive model
- Models of Inheritance: 1) dominant (in hetero vs homozygous); 2) recessive (if heterozygote,
allele remains biologically silent - only in genotype -, does not produce protein); 3) additive =
allele increases trait proportionately; 2 differing proteins from different alleles share function (=
per-allele model) → epistasis = interaction of genes
Hardy-Weinberg - for 2 alleles (A, frequency p; and a, frequency q), a er 1 generation of random mating, the
Equilibrium genotype frequencies of AA, Aa and aa in the population will be pˆ2, 2pq, qˆ2; [p + q = 1] and [pˆ2
+ 2pq + qˆ2 = 1]
- deviations from HWE may be explained by: 1) inbreeding (not random mating); genetic dri
(population is limited and isolated); 3) migration; 4) new mutations (rare, bc in sufficiently large
population, 1 generation is enough to reach equilibrium); 5) selection (disadvantage for an allele,
e.g. leading to fetal death); or methodological problems with the genetic study itself
- pedigree studies: good for rare disease caused by rare mutation
Genetic Association Studies - candidate gene study: testing a priori hypothesis that specific genes (also with low allele
frequency) are associated with disease; hypothesis-driven VS
- genome-wide association study: test over 1 million genetic variants that are frequent in the
population (SNPs); needs large sample size; agnostic, needs replication studies to check for
spurious associations;
, - the 2 methods are not mutually exclusive; usually represent population-based investigations
((non)diseased are unrelated); or in family members with mutations (linkage analysis); variation to
a gene is linked to an outcome
Linkage Disequilibrium = nearby stretches of a genome (tend to be) inherited together as a unit, result in hyplotype
blocks; → makes it hard to know whether the specific SNP has causal correlation (that's
important for therapeutic interventions, but not necessarily for determining risk)
- in genetic association studies, one goal may be to determine whether SNP is causally
associated with outcome
GxE & Psychiatric - nature/ heredity vs nurture/ environment: inseparable, mostly result of interaction → 37% MDD,
Disorders 65-80% schizophrenia, 60-85% bipolar
candidate GxE studies - most studied: serotonin transporter, MAOA, dopamine receptor DRD4 and DRD2, COMT, and
BDNF (brain-derived neurotrophic factors); are associated w/ multiple disorders (comorbidity)
- 5-HTTLPR & depression: serotonin transporter; s-allele is for less serotonin transporter mRNA
transcription, moderates serotonergic response to stress, makes a difference if exposed to
maltreatment/ stress → + suicidality and depression; still not clear (some show no effect); also
BDNF x SLEs → matches monoamine hypothesis (serotonins, antidepressants ↑ these levels)
- MAOA & antisocial behavior: codes for monoamine oxidase A, breaks NTs; short allele → low
MAOA expression → higher NTs; short allele + traumatic childhood; overall consistent support,
but also moderated by gender (stronger in boys)
- COMT & schizophrenia: codes for enzyme that degrades catecholamine NTs; valine allele = +
dopamine degradation; Val allele (but not Met) + cannabis use → psychotic symptoms; Met allele
+ stress → psychosis (but only for clinical patients); but inconsistent results (study designs?)
- GABA-receptor & alcohol dependence: x lower parental monitoring/ peer group antisocial &
substance availability; antagonists of GABA-receptor are effective in treatment of dependence
- Diathesis-Stress: environ only leads to psychop. when w/ inherent (genetic, temperament)
vulnerability; if not vulnerable (or counterbalanced w/ protective factors) = resilient
- alternative: differential susceptibility = Diathesis-Stress (+ negative exposure = for worse) &
Vantage Sensitivity (+ positive exposure = for better) → resilience as in Diathesis-Stress = "low
susceptibility to environmental influence" and vulnerability = "high susceptibility" - it's about genetic
sensitivity rather than vulnerability, studies should also cover the positive influence of environ
- it's evolutionary-inspired: the gene variants have survived natural selection, so must provide
benefits that counteract negative effects of heightened vulnerability
- but not necessarily always environmental sensitivity, sometimes it's just vulnerability
- limitations of candidate GxE studies: requires strong biological hypothesis, but only so much
is known, so a publication bias and wrong gene candidates; behaviors are rather polygenic;
difficult to replicate findings (small sample size, false positives?; most have been conducted
from Diathesis-Stress model perspective → shi ing to genome-wide
genome-wide association - hypothesis-free, threshold for statistical significance is very high; SNP-based heritability
studies (GWAS) estimates are a lot lower than in twin-studies (e.g. 20% vs 80%) - does not include GxE →
discrepancies in studies!
- genome-wide environment interaction studies (GWEIS): only a few so far, needs even larger
sample size!, more statistically complex; tests G and GxE associations on outcomes (SNP by
SNP approach to GxE)
- polygenic score-X-environment interaction studies: test interaction btw polygenic score (PGS)
& environmental variable; e.g. effects of childhood trauma in MDD was greater for those with
higher PGS for MDD (but unclear); → do more longitudinal studies!!; do + extensive, objective
measures of environ
future directions - obtain better environ measures: more accurate, avoid self-reports; experience sampling method
(ESM) = less recall bias, mood effect, App; consider multiple factors: childhood maltreatment,
season of birth, vitamin D, urbanicity, minority status, SES, cannabis use, SLEs in adulthood
- focus on transdiagnostic phenotypes: effects of GxE are transdiagnostic, comorbidity,
multifinality; e.g. 5-HTTLPR S-allele + daily stress → development of transdiagnostic emotion
dysregulation phenotype (no specific psychiatric disorder)
- developmental perspective: GxE depends on timing (more in childhood), do life course study
(gene-x-environ in childhood-x-environ in adulthood)
- new analytical approaches & study designs: not necessarily mutually exclusive, first consider
type of GxE, then the manner it operates
- focus on vulnerability genes (showing effect in GxE), instead of multiple-testing with all others
- for differential susceptibility genes, use MZ twins, see their difference (from non-shared
Lec 1 - Introduction - genetic sensitivity RA & depression, structural brain changes, immune system, serotonin &
depression → an interaction between processes, multidimensional intervention
application - allergy: early manifestation of immune dysregulation, classic psychosomatic disorder, complex
interactions, socially relevant; higher rate within families (35-75% genes), environ: too quick
incidence to be evolution alone, influence of stress, epigenetics processes: immune cell
differentiation & regulation; time window & dose influences, heterogeneous
- schizophrenia: heritability gap (twin studies overestimate heredity bc also share environ;
(20-35% genes)
Down's: 100% genes; MDD: 20-25% genes (a er correction for shared environ)
- clinical: change environ; Crisp-CAS9 changes DNA and alters epigenetics in many ways
How to Use an Article - pharmacogenomics = genetics to improve diagnosis and therapy
about Genetic Association - pedigree = diagram w/ heritable traits across 2+ generations of a family
- allele = variant of a gene; haplotype = alleles occurring together bc SNPs are close (so are
inherited together = genetic linkage); variant allele = the last frequent one (VS wild-type allele)
- cell → nucleus → chromosomes (22 pairs + 1 sex, winded by histones = proteins, to fit) → DNA
(bars are sugar-phosphate) → genes (a sequence of letters that encode for protein: adenine,
thymine, guanine, cytosine; each base pair is a rung; nucleotide = 1 base + phosphate groups; starts
w/ promotor, ends with terminator region)
- genome: DNA set of a human, 23 chromosomes egg + 23 chromosomes sperm
- phenotype: GxE, outward charac, personality (outside gene), genotype: genetic make-up
- SNPs = single-nucleotide polymorphism (single base pair change, vs common sequence);
synonymous SNP = SNP does not lead to change in amino acid
- VNTRs = variable number of tandem repeats
- epigenome: what is actually activated/ deactivated, some inprints survive through generations,
is changeable, depends on timing, changes constantly throughout the day!;
- epigenetics: without changes in DNA sequence, biochemical processes that switch genes
on/off, triggered by environ factors, signals come from the inside cell, nearby cells, and exterior;
modifies histones and accessibility to DNA transcription
- isoform = variant in amino acid sequence of a protein, the different proteins
Mechanisms of Genetics - gene to protein: DNA is transcribed into mRNA (messenger ribonucleic acid) → mRNA translated
into protein, in the ribosome (protein synthesis machinery, outside nucleus); first it's unwinded
genetic variation - >99% of DNA is similar across humans, <1% differs across individuals
- mutations = rare permanent change in base (deletion/ duplication/ DNA not repaired), <1%
- polymorphisms = 2+ variants of a gene (allele) occurring w/ at least 1% of the less common
variant; transmitted through generations, no major defect in bio functions: 1) presence/ deletion
of a stretch of DNA (involves DNA duplication = copy number variation = CNV); 2) repeating
patterns of DNA; 3) single-base pair change (SNPs), may be from segment that codes for protein,
or just influence cell function, the most common
- e.g. APOE gene: 3 alleles = e2, e3 (most common in white population), e4; additive model
- Models of Inheritance: 1) dominant (in hetero vs homozygous); 2) recessive (if heterozygote,
allele remains biologically silent - only in genotype -, does not produce protein); 3) additive =
allele increases trait proportionately; 2 differing proteins from different alleles share function (=
per-allele model) → epistasis = interaction of genes
Hardy-Weinberg - for 2 alleles (A, frequency p; and a, frequency q), a er 1 generation of random mating, the
Equilibrium genotype frequencies of AA, Aa and aa in the population will be pˆ2, 2pq, qˆ2; [p + q = 1] and [pˆ2
+ 2pq + qˆ2 = 1]
- deviations from HWE may be explained by: 1) inbreeding (not random mating); genetic dri
(population is limited and isolated); 3) migration; 4) new mutations (rare, bc in sufficiently large
population, 1 generation is enough to reach equilibrium); 5) selection (disadvantage for an allele,
e.g. leading to fetal death); or methodological problems with the genetic study itself
- pedigree studies: good for rare disease caused by rare mutation
Genetic Association Studies - candidate gene study: testing a priori hypothesis that specific genes (also with low allele
frequency) are associated with disease; hypothesis-driven VS
- genome-wide association study: test over 1 million genetic variants that are frequent in the
population (SNPs); needs large sample size; agnostic, needs replication studies to check for
spurious associations;
, - the 2 methods are not mutually exclusive; usually represent population-based investigations
((non)diseased are unrelated); or in family members with mutations (linkage analysis); variation to
a gene is linked to an outcome
Linkage Disequilibrium = nearby stretches of a genome (tend to be) inherited together as a unit, result in hyplotype
blocks; → makes it hard to know whether the specific SNP has causal correlation (that's
important for therapeutic interventions, but not necessarily for determining risk)
- in genetic association studies, one goal may be to determine whether SNP is causally
associated with outcome
GxE & Psychiatric - nature/ heredity vs nurture/ environment: inseparable, mostly result of interaction → 37% MDD,
Disorders 65-80% schizophrenia, 60-85% bipolar
candidate GxE studies - most studied: serotonin transporter, MAOA, dopamine receptor DRD4 and DRD2, COMT, and
BDNF (brain-derived neurotrophic factors); are associated w/ multiple disorders (comorbidity)
- 5-HTTLPR & depression: serotonin transporter; s-allele is for less serotonin transporter mRNA
transcription, moderates serotonergic response to stress, makes a difference if exposed to
maltreatment/ stress → + suicidality and depression; still not clear (some show no effect); also
BDNF x SLEs → matches monoamine hypothesis (serotonins, antidepressants ↑ these levels)
- MAOA & antisocial behavior: codes for monoamine oxidase A, breaks NTs; short allele → low
MAOA expression → higher NTs; short allele + traumatic childhood; overall consistent support,
but also moderated by gender (stronger in boys)
- COMT & schizophrenia: codes for enzyme that degrades catecholamine NTs; valine allele = +
dopamine degradation; Val allele (but not Met) + cannabis use → psychotic symptoms; Met allele
+ stress → psychosis (but only for clinical patients); but inconsistent results (study designs?)
- GABA-receptor & alcohol dependence: x lower parental monitoring/ peer group antisocial &
substance availability; antagonists of GABA-receptor are effective in treatment of dependence
- Diathesis-Stress: environ only leads to psychop. when w/ inherent (genetic, temperament)
vulnerability; if not vulnerable (or counterbalanced w/ protective factors) = resilient
- alternative: differential susceptibility = Diathesis-Stress (+ negative exposure = for worse) &
Vantage Sensitivity (+ positive exposure = for better) → resilience as in Diathesis-Stress = "low
susceptibility to environmental influence" and vulnerability = "high susceptibility" - it's about genetic
sensitivity rather than vulnerability, studies should also cover the positive influence of environ
- it's evolutionary-inspired: the gene variants have survived natural selection, so must provide
benefits that counteract negative effects of heightened vulnerability
- but not necessarily always environmental sensitivity, sometimes it's just vulnerability
- limitations of candidate GxE studies: requires strong biological hypothesis, but only so much
is known, so a publication bias and wrong gene candidates; behaviors are rather polygenic;
difficult to replicate findings (small sample size, false positives?; most have been conducted
from Diathesis-Stress model perspective → shi ing to genome-wide
genome-wide association - hypothesis-free, threshold for statistical significance is very high; SNP-based heritability
studies (GWAS) estimates are a lot lower than in twin-studies (e.g. 20% vs 80%) - does not include GxE →
discrepancies in studies!
- genome-wide environment interaction studies (GWEIS): only a few so far, needs even larger
sample size!, more statistically complex; tests G and GxE associations on outcomes (SNP by
SNP approach to GxE)
- polygenic score-X-environment interaction studies: test interaction btw polygenic score (PGS)
& environmental variable; e.g. effects of childhood trauma in MDD was greater for those with
higher PGS for MDD (but unclear); → do more longitudinal studies!!; do + extensive, objective
measures of environ
future directions - obtain better environ measures: more accurate, avoid self-reports; experience sampling method
(ESM) = less recall bias, mood effect, App; consider multiple factors: childhood maltreatment,
season of birth, vitamin D, urbanicity, minority status, SES, cannabis use, SLEs in adulthood
- focus on transdiagnostic phenotypes: effects of GxE are transdiagnostic, comorbidity,
multifinality; e.g. 5-HTTLPR S-allele + daily stress → development of transdiagnostic emotion
dysregulation phenotype (no specific psychiatric disorder)
- developmental perspective: GxE depends on timing (more in childhood), do life course study
(gene-x-environ in childhood-x-environ in adulthood)
- new analytical approaches & study designs: not necessarily mutually exclusive, first consider
type of GxE, then the manner it operates
- focus on vulnerability genes (showing effect in GxE), instead of multiple-testing with all others
- for differential susceptibility genes, use MZ twins, see their difference (from non-shared
