From Iusmgenetics

Contents 1 Genetic Mapping of Complex Disease 1.1 Complex Disease 1.2 Is it Genetic? 1.2.1 Twin Studies 1.2.2 Familial Aggregation 1.2.3 Mendelian forms of disease 1.2.4 Animal models 1.3 Genetics of Complex Disease 1.3.1 Multifactorial Liability Threshold Model 1.4 Risk Factor Concepts 1.4.1 Risk factor 1: Sex differences 1.4.2 Risk Factor 2: Degree of Relationship 1.4.3 Risk factor 3: severity of the defect 1.4.4 Risk Factor 4: Number of Affected Individuals 1.5 Incidence vs. Prevalence 1.6 Counseling in Multifactorial Traits 1.7 New types of genetic data 1.7.1 (Genetically) Personalized Medicine 1.7.2 “Missing heritability”

[edit] Genetic Mapping of Complex Disease

[edit] Complex Disease

• Both genes and environmental factors contribute to disease risk
  • Rarely can we strongly predict manifestation in an individual
• Genetic diseases seen most often in clinical practice
• Few genes and few environmental factors have been identified for these conditions
• Active area of research because these diseases consume a huge amount of resources.

• Examples of Complex Disease:
  • Atherosclerosis
  • Multiple Sclerosis
  • Diabetes mellitus
  • Alzheimer Disease
  • Epilepsy
  • Alcoholism
  • Dementia
  • Oral Clefts
  • Schizophrenia
  • Inflammatory Bowel Disease
  • Neural Tube Defects
  • Skeletal Disorders
  • Hypertension
  • Lung Cancer
  • Obesity
  • Parkinson Disease
  • Breast Cancer
  • Bipolar (manic depressive)
  • Depression

[edit] Is it Genetic?

• Single gene (Mendelian) disorders
  • obvious they are genetic
  • reviewing pedigrees makes the mode of inheritance obvious
    • Autosomal recessive (Phenylketonuria)
    • Autosomal dominant (Huntington Disease)
    • X-linked recessive (Duchenne muscular dystrophy)

• Other disorders, it appears that genetics is important
• But, there is NO recognizable pattern of inheritance (as checked for in pedigree)

• How do we quantify the genetic-ness of a disease?
  • Twin Studies
  • Familial Aggregation
  • Mendelian Form of Disease
  • Animal Models

[edit] Twin Studies

• Compare Monozygotic and Dizygotic Twins
  • Monozygotic Twins: genetically identical
  • Dizygotic Twins: like siblings (1/2 genome shared)
• Compare concordance rates of MZ and DZ twins

[edit] Familial Aggregation

• Familial aggregation is used when we can't get enough twin info.
• Increased risk for disease among family members of an affected individual
• Compare frequency of disease among first degree relatives of affected individuals with the frequency of the disease in the general population.
• Bipolar, as an example, has a RR (relative risk) of 4 or 5.
  • This is about as genetic as "complex disorders" get.
  • By comparison, Huntington disease (a single-gene disorder) has a RR of 10,000.

[edit] Mendelian forms of disease

• Some diseases have a subset of cases that can be shown to be related to one or few genes (and therefore it is demonstrated that the disease state has a genetic component).
• Breast Cancer
  • BRCA1, BRCA2
• Alzheimer Disease
  • Amyloid precursor protein, presenilin 1 and 2
• Parkinson Disease
  • alpha synuclein
  • parkin (autosomal recessive)

[edit] Animal models

• Identification of inbred animal strains with consistent findings of disease characteristics can be a useful predictor that a trait is genetic.
  • Can then perform experiments with controlled environment and planned mating to identify genes
• Hypertension
  • Rat lines that are stroke resistant and stroke prone

[edit] Genetics of Complex Disease

• Polygenic: determined by multiple genes
  • No environmental factors known.
• Multifactorial: determined by both genes and environmental factors

• As an example, height is both polygenic and multifactorial:
  • In 1850: nutrtion (an environmental factor) was the most prominent player in determining height.
  • In 2011: genes are the most prominent player (where food is abundant)

[edit] Multifactorial Liability Threshold Model

• Know that the point of this material is to be able to answer questions about an individuals increase or decrease in RR based on genetic and environmental information.

• Theoretical model designed to provide a means to explain gene interactions as well as possible environmental interaction with disease susceptibility
• Has not been proven to be true, but predictions of model are consistent with observations from clinical patients and their families

• Genetic liability: an individual's inherent (genetic) risk for a disease.
  • High genetic liability means you are (inhrently) closer to the threshold.
• All multifactorial diseases we discus are binomial in nature: you either have it or you don't.
• The threshold is the cutoff for the binomial state (having the disease or not).

• How many genes must be involved for a disease to manifest as multifactorial (binomial with a bell curve)?
  • Only two, really.
  • Even with only two genes segregating independently, we see bell curve distributions in genotype.

• Hypothesize that only those individuals who have inherited a sufficient number of susceptibility alleles at various genes will develop disease.
• Postulate that milder forms of the disease might be due to smaller number of susceptibility alleles.

[edit] Risk Factor Concepts

• There are four types of risk factors: sex, degree of relationship (to relation with disease state), severity of defect, number of individuals in pedigree.
• These risks represent the type of information one would collect from patients to assess their risk for a given complex disease.

[edit] Risk factor 1: Sex differences

• Some disorders have differing frequencies among the genders
  • individuals of the more rarely affected sex would need more susceptibility alleles to manifest disease
  • individuals of the more commonly affected sex would need fewer susceptibility alleles to manifest disease

• For example:
  • This model females are less often affected
  • Consider relatives of an affected male, the liability is shifted to the right, therefore increased risk of both affected male and female relatives
    • The liability for the whole family is shifted to the right; they are no longer in the normal population (because of their affected member).
  • Consider relatives of an affected female, the liability is shifted even further to the right, therefore even more increased risk of both affected males and female relatives
  • More area (of the curve) above the threshold (which does not move) indicates more people of the shifted population will be affected (both male and female, though one gender may have a higher threshold and therefore a lower percentage of infected members).

[edit] Risk Factor 2: Degree of Relationship

• First Degree: Share 1/2 of genes
  • parents, children, siblings
  • one line connects first degree relatives in a pedigree
• Second Degree: Share 1/4 of genes
  • uncles / aunts, nieces / nephews, g’parent, g’child, half-siblings
  • two lines connect second degree relatives in a pedigree
• Third Degree: Share 1/8 of genes
  • 1st cousins, g-g’parent, g-g’child, half-uncle / aunts, half-nephew / nieces
  • three lines connect third degree relatives in a pedigree

• Risk drops for 2nd degree relatives
  • It is expected that more closely related individuals would share more alleles in common than more distantly related relatives
  • Individuals more closely related to an affected individual are at higher risk of also being affected
  • The relative risk to a proband drops quickly as the affected relative moves from first degree to second, and on to third degree.

• Recall that on disease distribution diagrams, the threshold does not change but the curve is shifted from the general population position to represent the proband's population.
• When a first degree relative is affected, the proband's risk curve is shifted to the right.
  • A second or third degree affected relative would shift the curve to the right, too, but less to the right.
• Use this concept of shifting the curve to the right or left to consider risk in an individual.

[edit] Risk factor 3: severity of the defect

• Recurrence risk increases if the trait is more severe
  • More severe phenotype requires more susceptibility alleles
  • Greater risk for any spectrum of the phenotype (mild to severe)
• That is, we assume that if we observe a severe phenotype there is more susceptibility related genetic change underlying the phenotype.
  • Therefore, a more severe phenotype means an increased risk (as compared to another, less severe phenotype).

• Cleft Lip and Palate:
  • 60-80% of affected are male
  • Different causes:
    • isolated single gene (Mendelian inheritance)
    • part of a syndrome
    • teratogen
    • several genes/environment (Multifactorial)
• After these factors are ruled out, clefts are a function of many factors, including race.

• Different thresholds for different levels of disease severity
  • Consider relatives of a less severely affected individual, the liability is shifted to the right, therefore increased risk of both males and female relatives to be affected with any form of disease (mild to severe).
  • Consider relatives of a more severely affected individual, the liability is shifted even further to the right, therefore even more increased risk of both males and female relatives to be affected with any form of disease (mild to severe)
• Shift the proband's population risk farther to the right with increased severity of defect in affected relative.

[edit] Risk Factor 4: Number of Affected Individuals

• If there is more than 1 affected individual in the pedigree:
  • recurrence risk increases
  • suggests that the family might have even more susceptibility alleles segregating
• Be sure to also consider whether the family might have a single gene disorder
  • any other dysmorphism suggesting a syndrome?

[edit] Incidence vs. Prevalence

• Incidence: Proportion of individuals who have a disorder at birth.
  • That is: the proportion of the general population who get the disease.
• Prevalence: Number of individuals with a disorder in a given population at a particular time.
  • That is, the number of individuals of a particular population at a particular time.

• It is important to know incidence and prevalence of a disease because our default risk in predicting recurrence risk in a pt with an affected relative is the square root of the population incidence.

[edit] Counseling in Multifactorial Traits

• Recurrence risk for 1st degree relatives of an affected individual is square root of population incidence
• Example: Ventricular septal defect:
  • population incidence = 1/575
  • square root of 575 = ~24
  • recurrence risk for 1st degree relatives = 1/24
  • convert percentages to decimals or fractions!
    • Don't take the square root of the percent! (one way to represent incidence)
    • If population incidence is 2%, convert it to 2/100 = 1/50 and then take the square root of 50 such that the default relative risk is 1/7.

[edit] New types of genetic data

• Genome-wide SNP markers:
  • ~1 million SNPs to tag all common variants in an individual’s genome (~400 USD)
• Sequence data (whole-exome or -genome)
  • Identify all 60,000-100,000 rare and common variants in an individual’s genome
  • Rapidly becoming feasible (“$1,000 genome”)
  • Sequencing error rates low with current technology, but large numbers of false variants remain in any one person’s sequence

[edit] (Genetically) Personalized Medicine

• Challenge is interpretation
• We expect to identify both:
  • Rare variants that explain disease in a particular person / family
    • But it is hard to know which of these 90K are causing the disease.
  • Variants that increase susceptibility to disease across an ethnic group or larger population
    • But it is hard to know which have real consequence in terms of treatment and correction.
• How to translate this to prognosis, lifestyle modification recommendations, etc. in any one individual??

[edit] “Missing heritability”

• GWAS studies typically detect common variants explaining <20% (combined) of total variation
  • GWAS is capable of associating SNPs with variability but it doesn't always explain inheritiablity of the complex disease.
• The rest of the difference (non SNP difference) is probably explained by:
  • Gene-gene interactions (epistasis)
  • Gene-environment interactions
• Only a very small proportion of variation in trait or disease risk is explained by combined effects of common SNPs.