Exam 4! YAY SLAY THE DAY

Genetic variation

Differences in DNA content or structure among individuals.

Human DNA similarity

Any two humans share approximately 99.5% identical DNA.

Human haploid genome

23 chromosomes containing approximately 3.1 billion nucleotides.

Known human genetic variants

More than 100 million variants have been identified.

SNP

Single-nucleotide polymorphism; a difference in one DNA nucleotide.

INDEL

Insertion or deletion of DNA nucleotides.

Structural variant (SV)

A large block of DNA that is duplicated, deleted, inverted, or rearranged.

1000 Genomes Project

Population-scale project cataloging human genetic variation.

TOPMed

Trans-Omics for Precision Medicine; study with about 180,000 participants in the review.

UK Biobank

Large genetic study with about 500,000 volunteers in the review.

All of Us Research Program

Population research program with about 790,000 participants in the review.

SNP calling

Identifying a polymorphic nucleotide position.

Genotype calling

Assigning an individual's genotype at a genetic site.

Reference allele

Allele represented in the reference genome.

Non-reference allele

Allele differing from the reference genome.

Old SNP-calling method

Count allele reads and apply a frequency threshold.

Old SNP-calling threshold

20–80% non-reference allele reads in the review example.

High read coverage

More than 20 reads at a site in the older approach.

Modern SNP calling

Probabilistic method that incorporates sequencing uncertainty.

Site frequency spectrum (SFS)

Distribution of allele frequencies across genetic variants.

Rare variant

Genetic variant occurring at low frequency in a population.

Main SFS observation

Most genetic variants are rare.

Recent population growth

Associated with an excess of rare human genetic variants.

Linkage disequilibrium (LD)

Nonrandom association of alleles at different genetic sites.

LD haplotype block

Cluster of linked SNPs with strong linkage disequilibrium.

Haplotype

Combination of alleles inherited together along a chromosome.

Recombination hotspot

Region with elevated recombination that can define LD block boundaries.

Haplotype association advantage

Using linked SNPs together can be more robust than using individual SNPs.

Hardy-Weinberg equilibrium (HWE)

Expected genotype frequencies from allele frequencies under equilibrium assumptions.

p

Frequency of one allele.

q

Frequency of the other allele.

p + q = 1

Sum of the two allele frequencies.

p²

Expected frequency of homozygotes for the p allele.

2pq

Expected frequency of heterozygotes.

q²

Expected frequency of homozygotes for the q allele.

p² + 2pq + q² = 1

Hardy-Weinberg genotype frequency equation.

HWE example p = 0.1, q = 0.9

AA = 0.01; AG = 0.18; GG = 0.81.

Expected AG among 100 people

18 heterozygotes when p = 0.1 and q = 0.9.

Expected AA among 100 people

1 homozygote when p = 0.1.

Structural variation size

Genomic rearrangements affecting more than 1 kb in the review.

Duplication

A DNA segment is copied, increasing its number of copies.

Deletion

A DNA segment is removed.

Inversion

A DNA segment reverses orientation.

Translocation

A DNA segment moves to another chromosomal location.

Loss of heterozygosity (LOH)

Loss of one allele's contribution in a previously heterozygous region; can occur in cancer.

Copy number variation (CNV)

Variation in the number of copies of a DNA segment.

CNV mechanisms

Large DNA deletions or duplications.

Germline CNV

Inherited or reproductive-line copy number variation.

Somatic CNV

Copy number change acquired in body cells, such as tumor cells.

CNVs and cancer

Tumors can contain acquired DNA gains and losses.

Whole-exome sequencing (WES)

Sequencing the protein-coding regions of the genome.

Exome

Collection of protein-coding exons.

Exome genome fraction

Approximately 1% of the genome.

WES advantage

Cheaper and faster than whole-genome sequencing.

WES limitation

Does not comprehensively sequence noncoding genomic regions.

Whole-genome sequencing (WGS)

Sequencing DNA across the entire genome.

Read depth

Number of sequencing reads covering a genomic region.

Read depth matrix

Read coverage measurements across exons and samples.

Sample-specific bias

Read-depth differences caused by characteristics of samples.

Library size bias

Different total sequencing read counts among samples.

Batch effect

Technical variation caused by different processing batches.

Population stratification

Genetic differences associated with population ancestry.

Exon-specific bias

Read-depth differences caused by characteristics of exons.

Exon length bias

Coverage differences related to exon length.

GC content bias

Sequencing coverage differences related to GC percentage.

Exon capture efficiency

How effectively an exon is enriched during sequencing preparation.

Amplification efficiency

How efficiently DNA fragments are copied during PCR.

Latent factors

Unobserved influences contributing to read-depth variation.

CNV hotspot

Genomic region where copy number changes recur.

Neuropsychiatric CNV associations

Certain deletions and duplications are enriched in neurological or psychiatric disorders.

Reciprocal CNVs

Deletion and duplication affecting the same genomic region.

GWAS

Genome-wide association study testing genetic variants for trait associations.

Manhattan plot

Graph showing genome-wide statistical association signals.

Manhattan plot X-axis

Chromosomal or genomic position.

Manhattan plot Y-axis

Negative log10 of the association p-value.

Manhattan plot peak

Genomic region with a strong statistical association.

Significance threshold

Cutoff used to identify statistically significant associations.

Higher −log10(p)

Smaller p-value and stronger statistical significance.

GWAS outputs

Association p-values and effect sizes with directions.

Effect size

Estimated magnitude of a variant's association with a trait.

Effect direction

Whether the associated allele increases or decreases a measured outcome.

GWAS tag SNP

Representative SNP associated with a genomic signal.

GWAS hits are mostly noncoding

Most associated variants lie outside protein-coding sequences.

Noncoding variant

Genetic variant outside protein-coding sequence.

eQTL

Expression quantitative trait locus; genetic variant associated with gene expression.

eQTL mapping

Testing whether genotype predicts gene expression.

Minor allele

Less frequent allele at a genetic site.

Minor allele dosage

Number of minor alleles, coded 0, 1, or 2.

AA genotype coding

0 when A is major and G is minor.

AG genotype coding

1 when A is major and G is minor.

GG genotype coding

2 when A is major and G is minor.

eQTL regression

Y = μ + βx + ε.

Y in eQTL regression

Normalized gene expression.

μ in eQTL regression

Mean or baseline gene expression.

x in eQTL regression

SNP genotype dosage.

β in eQTL regression

Regression coefficient representing effect size.

ε in eQTL regression

Residual error.

eQTL null hypothesis

β = 0; genotype has no linear association with expression.

eQTL statistical test

t-test of whether β differs from zero.

Unique molecular identifier (UMI)

Molecular barcode used to identify original molecules.