Psy 5137 - Exam 2: Molecular Genetic Concepts and Methods I and II

2 lectures

What is linkage analysis?

A method that searches for chromosomal segments that co-segregate with a disease phenotype through families, based on the fact that genes physically close on a chromosome remain linked during meiosis.

What are limitations of linkage analysis?

Requires large, informative families; does not identify causal variants (only regions harboring them); and is better suited for traits with moderate to large effects rather than small effects.

What is positional cloning?

A method of gene identification where a gene is identified by its approximate chromosomal location (candidate region) but not its specific function.

Can positional cloning identify specific genetic variants or causes?

No, it only narrows down the genomic region, it cannot directly identify causal variants or their functions.

What is allelic association?

A population-level (not family-based) association between allele status and phenotype in case-control or population studies.

What metric is used to represent risk in allelic association studies?

Odds ratio (OR)

• OR > 2 indicates strong association

• OR ≈ 1 indicates no association.

What genetic markers are typically used in allelic association studies?

Single nucleotide polymorphisms (SNPs).

What is the candidate gene approach?

Researchers select specific genes based on prior knowledge of their biological relevance to a trait and test whether SNPs in those genes are associated with the phenotype.

What were the strengths of the candidate gene approach?

High statistical power for small effects (with large sample size) and ability to test specific biological pathways for involvement in a disease.

What were the weaknesses of the candidate gene approach?

It required predefined hypotheses, was prone to false positives due to population stratification, publication bias, and p-hacking, and often failed to replicate findings.

What ultimately came of the candidate gene approach?

It was largely replaced by genome-wide association studies (GWAS), which allow hypothesis-free exploration across the whole genome.

What is SNP imputation?

A method for estimating missing genetic data by comparing known SNPs with a reference panel to predict unobserved variants.

What is SNP tagging?

A strategy that uses linkage disequilibrium (LD) to predict nearby SNPs—SNPs that are close together tend to be correlated and can serve as proxies for each other.

How do imputation and tagging take advantage of linkage disequilibrium (LD)?

They use the fact that nearby SNPs are inherited together; knowing one SNP allows prediction of nearby ones without directly genotyping them.

What is linkage disequilibrium (LD)?

The non-random association between alleles at linked loci; alleles close together on a chromosome are often inherited together.

What is a haplotype and how is it related to LD?

A haplotype is a set of alleles at linked loci inherited together on the same chromosome; LD occurs because haplotypes persist over generations until recombination breaks them apart.

How do meiosis, recombination, and generations affect LD?

LD decays over time with recombination; the closer two loci are, the less likely recombination will separate them. The older a mutation, the smaller the region of LD that remains.

What factors influence the extent of LD?

Physical distance between loci, number of generations since the mutation arose, recombination frequency, mutation rate, population size, and selection.

What is the basic idea of a genome-wide association study (GWAS)?

A hypothesis-free method that scans the entire genome to find common SNPs associated with traits or diseases.

Are GWAS designed to detect common or rare variants? Why?

Common variants (frequency >1%), because rare variants are too infrequent to provide adequate statistical power in typical sample sizes.

Why are common variants more likely to be detected in GWAS?

They tend to be ancient, shared across populations, and in LD with nearby markers that can be measured in large samples.

How does GWAS sample size affect the number of detected variants?

Larger sample sizes increase statistical power, allowing detection of more variants, especially those with small effects.

How does sample size affect total variance explained in a GWAS?

Larger samples detect more variants, collectively explaining a greater proportion of the phenotype’s variance.

How much variance is typically explained by each significant SNP?

Each SNP usually explains a very small fraction of total variance; many SNPs together contribute to complex traits.

What did Sanders et al. (2008) find in their candidate gene study of schizophrenia?

They found 30 variants significant at p < .05 and 3 at p < .01, but none reached genome-wide significance given the large number of tests.

Why did the Sanders et al. (2008) findings not reach genome-wide significance?

Either the study was underpowered (true effects too small to detect) or the chosen candidate genes were false positives—most likely the latter.

….

….

What is meant by genetic architecture of complex/multifactorial traits?

Genetic architecture refers to the characterization of the specific genetic variants that influence complex traits, including:

• The number of contributing variants

• Their frequency in the population

• The magnitude of their effects on the phenotype

• The mechanisms by which they affect risk

For any complex phenotype, there are likely thousands of genetic variants contributing to risk, ranging from very rare to very common.

What range of effect sizes do common variants identified in a GWAS typically have on the phenotype? What about rare variants?

• Common variants are likely to have a small to very small phenotypic effect. Likely to be regulatory rather than change the protein product, and will require very large samples to detect.

• Rare variants can have small to large phenotypic effects, but only large effect variants are detectable. Large effect variants are likely to change protein product, and will require special (e.g. exome sequencing) strategies to detect.

In terms of GWAS, what do geneticists typically predict will be the relationship between effect size and allele frequency of detected hits?

What do we predict about common variants of large effect?

What about rare variants of small effect?

• Common variants with high allele frequencies are more likely to be detected in GWAS, but typically have small effect sizes.

• Rare variants with low frequencies can have larger effects, but are less likely to be detected.

• Common variants of large effect are rare due to evolutionary pressure.

• Very rare variants of small effect are nearly impossible to detect without extremely large samples.

What is meant by the common disease common variant model and the common disease rare variant model?

• CDCV (Common Disease–Common Variant): Many common variants of small effect → fits GWAS and reflects reality best.

• CDRV (Common Disease–Rare Variant): Many rare variants of large effect → each case may have a unique mutation.

  • GWAS is based on CDCV model.

Which do common vs rare variants affect?

• Common variants: Affect gene expression (regulatory changes).

• Rare variants: Affect gene product (protein structure/function).

What is missing heritability?

The gap between heritability estimated from family/twin studies and that explained by GWAS.

3 causes of missing heritability:

• Inflated twin estimates

• Non-additive genetic effects

• GWAS underpowered to detect tiny SNP effects (most likely option)

Between increasing GWAS sample sizes, SNP heritability estimates, and inclusion of rare variants, is it feasible to account for most of the so-called missing heritability?

No, even though larger GWAS and inclusion of rare variants can explain more variance, it is unlikely that all missing heritability will be accounted for.

When it comes to GWAS, how do we typically quantify the proportion of genetic variance accounted for in a phenotype?

Genetic variance is quantified using a Polygenic Risk Score (PRS).

How is a polygenic risk score computed?

A PRS is a weighted linear combination of SNPs identified in GWAS, with each SNP weighted by its strength of association with the phenotype.

What information is captured in a polygenic score?

• The PRS captures an individual’s overall genetic risk for a trait or disease.

• Those with the disorder tend to have higher PRS than those without it.

What is SNP heritability?

The total phenotypic variance explained by all SNPs included in a GWAS (a theoretical estimate assuming infinite sample size).

Pleiotropy

One gene affects multiple traits.

Genetic correlation

Measures overlap in genetic effects between two traits.

• Example: Schizophrenia and bipolar disorder show a genetic correlation > .60, suggesting shared biological roots and blurred diagnostic boundaries.

Why is it difficult to identify causal mechanisms from GWAS hits?

Because common variants have tiny effects and are often in linkage disequilibrium (LD) with causal variants.

However:

• We can study whether GWAS hits cluster in specific gene sets or pathways (e.g., synaptic pruning genes in schizophrenia).

• This helps infer biological mechanisms even when individual SNPs’ roles are unclear.

How could GWAS be clinically useful in the future?

• Identify biological pathways for new drug targets.

• Use PRS to identify high-risk individuals for preventive or early interventions.