Genetics Final Exam PDF

Genetics Final Monday, December 9, 2024 11:31 PM 1. Genetic Architectures (Monogenic to Omnigenic) Monogenic Traits: Governed by a single gene; often follow Mendelian inheritance patterns. ○ SEGREGATION ○ Single genetic variant ○ Variants are rare ○ Example: Cyst...

Genetics Final Monday, December 9, 2024 11:31 PM 1. Genetic Architectures (Monogenic to Omnigenic) Monogenic Traits: Governed by a single gene; often follow Mendelian inheritance patterns. ○ SEGREGATION ○ Single genetic variant ○ Variants are rare ○ Example: Cystic fibrosis. Polygenic Traits: Controlled by multiple genes with additive effects. ○ AGGREGATION ○ Combined Many Variants ○ Small effect on risk ○ Often common variants in population ○ Have environmental factors ○ Example: Height. Architecture ○ The underlying genetic basis of phenotypic variation for a given trait or disease Oligogenic Model: Small amount of genes, but more than two can influence a trait or disease Omnigenic Model: Nearly all genes in a genome can influence a trait indirectly through core genes in relevant pathways. ○ Key Idea: Pervasive pleiotropy across the genome. Quantitative Traits lead to Qualitative: ○ LDL Genes (Quantitative) lead to Hypercholesterolemia (Disease is Qualitative) 2. Dichotomous Phenotypes Threshold Model: Liability for disease risk is a hypothetical quantitative trait. When liability crosses a certain threshold, disease occurs. ○ Burden of Genetic and Environmental risk factors ○ Example: Schizophrenia. Endophenotypes: Multiple measurable quantitative phenotypes, each with their own genetic architecture ○ Liability is an accumulation of unfavorable phenotypes ○ Not incompatible with the threshold model 3. Heritability and Variance Explained Heritability is the proportion of the total variance (NOT MEAN) in a quantitative trait that can be attributed to genetic variation ○ Heritability values greater than.3 for a trait are generally considered high enough to make it worthwhile to look for genes affecting the trait ○ Applies to POPULATIONS not individuals ○ It is NOT the probability that a trait is inherited or has a genetic basis Heritability Formula: Heritability = 2 x (MZ - DZ) ○ Changes in Components: ▪ If environmental variance decreases, heritability increases because less variance is attributed to non-genetic factors. Key Terms: ○ Additive Genetic Variance (VAV_AVA): Variation due to additive effects of alleles. ○ Environmental Variance (VEV_EVE): Variation due to environmental factors. Genetics Final Page 1 4. Genetic Association Studies Study designs for the genetics of complex diseases ○ Low variance with low effect size = difficult ○ Very low var with large effect size = Link Analysis or Clinic Exome Sequencing ○ Very high var with low effect = Association Study ○ Very high var with high effect = unlikely to exist (Or can use GWAS) ○ Turning SNP alleles into integer. (C = dominant) Screen clipping taken: 12/10/2024 1:30 PM For each genetic variant: 1. Choose an allele to be the coded allele 2. Code variant using additive model: 0, 1, or 2 copies of coded allele (dosage) 3. Calculate OR and p-value 4. Determine statistical significance using p. Candidate Gene Studies: Test specific genes hypothesized to be involved in a trait. ○ Hypothesis-based: a selection of genetics variants is made based on a hypothesis and ONLY this selection is examined for association with disease ○ Strengths: Targeted, lower cost, hypothesis driven ○ Weaknesses: Subject to publication bias, hard to use to discover new biology, we aren't very good at predicting which genes are likely to harbor associated variants, or which variants in a region are like to be important Genome-Wide Association Studies (GWAS): Assess millions of variants across the genome for associations with a trait, analysis for all measured genetic variants ○ Large number of variants ○ Hypothesis-free ○ No issue with publication bias (all results are reported) ○ Data is usually based on SNP arrays (so far) ○ Use basic genetic variant step from above ○ Cannot use Bonferroni correction (too conservative) ▪ GWAS are corrected for 1 million independent tests ▪ Genome-wide significance = < 5 x 10 ^ -8 ▪ Ensures false positives are minimum, but comes at the cost of statistical power ▪ Implication: with a given sample size the minimum detectable effect size is lower for candidate gene studies than for GWAS ○ GWAS NEEDS LARGE SAMPLE SIZES ▪ Investigators gain from combining results from across multiple studies ▪ Meta-analysis is used to combine studies ○ Replication ▪ Replication associations in an independent sample drastically reduces false positive rate ○ Power in GWAS ▪ Sample size plays a more critical role in GWAS than any other epi study ▪ # of cases is more important than # of controls □ Controls don't add power after 4 controls per case Genetics Final Page 2  □ Controls don't add power after 4 controls per case ▪ Power depends on minor allele frequency (MAF) and effect size ▪ Overall contribution of a variant on disease risk in a population depends on MAF and effect size ○ Confounding ▪ Population stratification ▪ Family structure ○ Genomic Inflation Factor ▪ If close to 1, there may be some associations, but mostly devoid of associations. (You want this) Larger than 1.1 is unacceptable inflation ○ Strengths: Unbiased discovery, identifies new loci. ○ Weaknesses: Requires large sample sizes, cannot confirm causality. Statistical Power Considerations: ○ Increases with effect size, allele frequency, and sample size. ○ Rare variants require larger sample sizes for detection. 5. Confounding in Genetic Studies Non-Causal Associations: ○ Chance: Even if there is no true association, sometimes false positive associations occur by chance ▪ Gold standard is Bonferroni correction. ▪ Create new threshold using alpha / (N hypotheses) (=N SNPs) ▪ Decreases statistical power ▪ Stricter alpha = lower power ▪ Increase power by increasing sample size ○ Hidden Population Stratification: Population may include several subsets, in some of which both the variant and the disease are more common (type of confounding) ▪ An issue when these conditions are met: □ Multiple populations / ancestry groups □ Recently admixed populations / ancestry groups □ The prevalence of the disease differs by population □ The allele frequencies of the genetic variants differ by population ▪ Family structure can also confound genetic association structure ▪ Variants (especially rare) differ in specific families compared to the general population ▪ Disease aggregates in families partially due to non-genetic factors (shared environment) □ Mitigate by stratify data by ancestry group □ Exclude related individuals ○ Linkage Disequilibrium (LD): Variant is linked to a measured or unmeasured causal variant ▪ Alleles at two loci are found together more or less often than expected by chance ▪ Non-random association of alleles at two loci ▪ Haplotypes do not occur at the expected frequences given the allele frequencies ▪ Use r-squared to determine measure □ R-squared = 1 ONLY if just 2 of the 4 possible haplotypes are observed (perfect disequilibrium) ▪ Tag SNPs □ If two SNPs are in strong LD, only one SNP is needed (Tag SNPs) □ Unlikely that a genetic association of the tag SNP with disease is causal □ More likely that one of the unmeasured variants the tag SNP is in LD with is causal □ Tag SNP does point to a causal association in the region 6. Linkage Disequilibrium, Haplotypes, and Imputation Genetics Final Page 3 6. Linkage Disequilibrium, Haplotypes, and Imputation Linkage Disequilibrium (LD): Non-random association of alleles at different loci. ○ Example: SNPs close on a chromosome. Haplotypes: Groups of alleles inherited together. Imputation: Predicting untyped genotypes based on known LD patterns. ○ Used to predict missing or unmeasured genotype values prior to analysis ○ Use knowledge of haplotypes combined with measured genotypes to inform this predication ○ This knowledge of haplotypes comes from "reference panels" based on whole genome sequencing ○ Reference Panels ▪ Whole genome sequencing datasets that provide information on haplotype frequencies ○ Perfectly imputed genotypes r-square = 1 ○ Imperfectly imputed genotypes have r-squared < 1 ○ Lower imputation quality = more error = lower power 7. Pharmacodynamics vs. Pharmacogenetics Pharmacodynamics: How a drug affects the body. ○ Varies by type of molecules or tissues affected, for example: ▪ Enzymes ▪ Receptors ▪ Ion Channels ▪ Transporters ○ The genetics of pharmacodynamics are often more complex than pharmacokinetics Pharmacokinetics: What body does to the drugs ○ How drug is processed by the body: ▪ Absorbed ▪ Transported to target tissues ▪ Metabolized ▪ Excreted ○ Most easily studied by measuring ▪ Drug concentrations in plasma or urine ▪ Drug metabolite concentrations in urine Pharmacogenetics (Friedrich Vogel 1959): How genetic variation affects drug response. ○ Aims to take the guesswork out of prescribing drugs by using genetic information to help identify those likely to respond to a drug without unwanted side effects ○ Overall Goal: To identify genetic variants that affect the action of a drug and help to target its use to those most likely to be helped by it and least likely to be harmed by it. ○ Example: Warfarin metabolism and VKORC1 gene variants. ○ Identification of genetic profiles that accurately predict a person's response to drugs is likely to increase the overall efficacy and safety of pharmaceuticals Identifying Pharmacogenetics Associations ○ GWAS on outcome among study participants taking the medication (efficacy) ○ GWAS on adverse reactions (toxicity) ○ GWAS on levels of the drug and/or its metabolites (efficacy and toxicity) ○ Genome-wide interaction study (GWIS) to test for interactions between variants and medication (efficacy) 8. From Genetic Associations to Drugs Pipeline Using GWAS: 1. Identify a causal gene for disease i. Functional studies are usually required: Gene silencing in relevant cell lines, Gene Genetics Final Page 4  i. Functional studies are usually required: Gene silencing in relevant cell lines, Gene knockout in animal models 2. Identify the direction of the effect of the gene i. Does coded allele of the SNP increase or decrease risk of disease ii. Does the coded allele cause a loss-of-function or a gain-of-function? 3. Determine the effect of the SNP on the disease. 4. Assess whether variants in the drug target have pleotropic effects (Optional) i. Whether they have effects on other phenotypes ii. This could predict adverse side effects for a drug 5. Series of studies for safety and efficacy: Preclinical trials and clinical trials (phase 1 to 3) i. Culminates in randomized controlled trials ii. FDA approval Drug Targets: Genes (or proteins) that should be stimulated or inhibited by a drug in order to treat disease 9. Genetic Risk Scores (GRS) A single variable representing the cumulative "burden" of risk alleles for a disease Unweighted GRS: Sum of risk alleles. Weighted GRS: Each allele is weighted by its effect size. (Multiply Log(OR) by allele score and get sum) Applications: Predict disease risk, stratify patients for interventions. 10. Gene Therapy Focus has been on severe Mendelian disorders Most gene therapy to date is largely based on gene replacement For Mendelian Diseases: CRISPR-Cas9 to correct mutations. ○ Changing specific base pairs to remove a harmful mutation or introduce a beneficial mutation For Complex Diseases: Emerging, but challenging due to polygenic nature. Example: Sickle cell disease gene editing. 11. Cancer Genetics Cancer Development: ○ Cancer: ▪ Not a single disease, but a set of disorders that involve abnormal cell proliferation ▪ Promotion of cell proliferation ▪ Inhibition of apoptosis (cell death) ▪ Cells show loss of cellular differentiation ▪ Invasiveness of surrounding tissue ▪ Propensity to spread to distant sites (metastasis) ▪ Inherently genetic disease, but MOST are NOT inherited. ▪ Most monoclonal - derived from a single mutated cell ○ Multiple-Hit Theory: Accumulation of mutations in oncogenes and tumor suppressor genes. Cancer Genes: ○ Proto-oncogenes: Promote cell growth (e.g., RAS). ▪ Gain of function mutations turn these into oncogenes promoting uncontrolled cell growth and proliferation ▪ Dominant at the cellular level (even cells with a single mutated copy can cause cancer) ▪ Dominant for individuals ▪ Germline mutations are rare ▪ What causes oncogenes: Genetics Final Page 5  ▪ What causes oncogenes: □ Over-express a normal protein □ Expressing an aberrant protein ▪ Mutation types that can produce oncogenes □ Missense mutations □ Gene duplications □ Chromosomal rearrangements or translocations ○ Tumor Suppressor Genes: Control the progression of a cell through the cell cycle ▪ Loss of function leads to mutation ▪ Recessive at the cell level (only cells with 2 mutated copies become cancers ▪ Dominant for individuals (heterozygotes will usually develop cancer) ▪ Mutation in a tumor suppressor gene □ Promoter mutations can block transcription □ Nonsense mutations or frameshifts produce no protein or a non-functional protein ▪ Aneuploidy or deletions in tumor cells □ Loss of whole chromosomes or regions with tumor suppressor genes may spur tumor progression □ (RB1 - retinal blastoma) ○ DNA Repair Genes: Detect and remove DNA damage or mutations ▪ Loss of function leads to increased mutation rate ▪ (Some count DND repair genes as tumor suppressor genes) ▪ Recessive at cell level, dominant for individuals ▪ Mutations IN At least 34 DNA repair genes can cause cancer 12. Precision Medicine Complex Diseases: Stratify patients by genetic profiles to improve treatments (e.g., diabetes). Cancer: Identify driver mutations for targeted therapies (e.g., EGFR inhibitors for lung cancer). ○ Cancer subtypes can be distinguished by morphological features ○ More accurate diagnosis allows for more tailored treatment ○ Knowledge of specific mutations involved in cancer facilitates drug discovery process ○ Patients that lack the specific mutation targeted by a drug will not benefit and can be harmed, so precision medicine is required. Personalized medicine: creation of drugs or medical devices that are unique to one patient Precision medicine: Applying the existing preventive or therapeutic interventions that are best suited according to the characteristics of the individual patient Genetics Final Page 6