Neurogenetics Flashcards
What is DNA made up of and what are the three main componenets?
Repeating nucleotide units made up of:
▪️ Phosphate group
▪️ Sugar
▪️ Nitrogenous base (ATGC)
How identical is the genetic code between people?
99.9%
What is genetic variation?
Differences in our DNA sequence
What are SNPs?
Single nucleotide polymorphisms - the most common type of variation that can lead to differences in DNA sequence (typically not deleterious/mutation)
E.g., a G swapped for a T
When are SNPs typically benign?
▪️ Between genes (non-coding areas)
▪️ Within genes but don’t change protein coding regions
▪️ Within genes and change protein sequence, but small effect and possibly advantageous to protein function
What are Mendelian diseases?
▪️ Direct relationship between gene and disease
▪️ Single gene disorder - change in DNA/gene will definitely lead to disease
▪️ Recognisable heritable patterns
▪️ Much less common (but much more significant effect)
What are the main examples of Mendelian diseases in neuropsychiatry?
▪️ Huntington’s disease
▪️ Familial Alzheimer’s disease
▪️ Motor Neurones Disease/FTD
▪️ Fragile X syndrome
▪️ Tuberous Sclerosis
▪️ Spinocerebellar ataxia
What are the majority of psychiatric and behavioural disorders?
Complex diseases
What are complex diseases?
▪️ Gene = increased risk but does not cause disease directly
▪️ May observe familial pattern but less clear
▪️ Disease is due to interaction between multiple genes or genes + environment
How do Mendelian forms of a disease typically differ from their complex counterparts?
Generally more severe and earlier onset of symptoms (e.g., Alzheimer’s)
What are the main types of large chromosome-level changes to DNA?
▪️ Deletion (e.g., 22q11.2 = VCFS, increased risk of SCZ and autism)
▪️ Duplication (e.g., trisomy 21 = DS, double APP = EOAD)
▪️ Rearrangement
What are the main types of smaller genetic mutations?
▪️ Deletion (e.g., PSEN1 and AD)
▪️ Duplications (e.g., CAG expansion in HD)
▪️ SNP substitutions (e.g., PSEN1 and AD)
How might a SNP substitution cause disease?
▪️ Changes coding of protein (missense mutation)
▪️ Changes processing of mRNA which changes protein
▪️ Changes regulation of gene expression
How many missense mutations in PSEN1 can cause AD?
~93
What is a dominant disease?
▪️ Only needs ONE mutant allele to get the disease
▪️ 50% chance of passing disease on
▪️ Never “skips” a generation
What is a recessive disease?
▪️ Needs TWO mutant alleles to get the disease
▪️ 25% chance of passing disease on
▪️ 50% will be carriers (one mutant allele) - asymptomatic
How common are SNPs?
▪️ Very!
▪️ ~10 million in human genome
▪️ Once every 300 nucleotides
What is the Hardy Weinberg equation for SNP frequency?
p2 + 2pq + q2 = 1
p = frequency of “A” allele
q = frequency of “a” allele
p2 = probability of homozygous AA
2pq = probability of heterozygous Aa
q2 = probability of homozygous aa
What is heritability?
The proportion of phenotypic variance attributable to additive genetic factors
2(rmz - rdz)
What is an MAF?
Minor Allele Frequency
What is a GRR?
Genotype Relative Risk
What are the main types of heritability?
▪️ Rare high risk mutations with clear impact (few highly penetrant genes)
▪️ Common low-risk genetic variants (MAF >2%, GRR = 1-2)
▪️ Rare moderate-risk genetic variants (MAF <2%, GRR = 2-50)
▪️ Rare low-risk genetic (MAF <2%, GRR = 1-2)
What factors contribute to genetic risk in a population?
▪️ Number of risk alleles
▪️ Frequency of each risk allele (major or minor)
▪️ Effect size of risk allele
▪️ Interaction between alleles (additive or multiplicative)
Why are diseases such as SCZ, influenced by multiple genes, not diagnosed as continuous traits?
Liability threshold model - certain threshold must be exceeded before disease occurs/is recognised
E.g., relatives have shifted distribution of trait reflecting increased liability (move distribution curve to the right)
How can we identify rare alleles that cause Mendelian disease (high effects)?
▪️ Linkage analysis
▪️ Candidate gene analysis
What can linkage analysis be used for?
▪️ Identifying genes for monogenic diseases
▪️ Identifying large genomic regions containing susceptibility genes
How does linkage analysis work?
▪️ SNPs throughout genome as markers to tag the genotypes of those with and without disease within a family
▪️ If tag genotype is physically close to the disease SNP, it will always be inherited by those with disease
▪️ Relies on recombination between chromosomes and alleles being inherited together
What are linked alleles?
Alleles that are inherited together more than 50% of the time (more than chance/expected)
(Often syntenic genes - on the same chromosome)
What are the main problems with linkage analysis?
▪️ Relies on large pedigrees from one family, especially to find very rare, low GRR genes
▪️ Smaller pedigrees may be heterogeneous (different genetic causes)
▪️ Incomplete penetrance?
▪️ Complex inheritance patterns
▪️ Marker linked with disease usually still along way away so may not find faulty gene
What can be used now to solve the problems with linkage analysis?
Genome Wide Association Studies
(can measure more precisely too)
What is a GWAS?
A technique that compares differences in the frequencies of specific alleles between diseased individuals and healthy controls from the population
Can be:
▪️ Case controlled
▪️ Family based
▪️ Quantitative trait study
Associated alleles can….
▪️ Directly influence risk
▪️ Be in linkage disequilibrium with the disease-causing allele (detecting susceptibility polymorphisms by proxy)
What is imputation?
Using statistics to calculate genotypes at “all” loci, filling in the gaps of SNPs that aren’t actually measures
What are the main limitations of GWAS?
▪️ Not good for rare variants (<5%)
▪️ Multiple testing with large number of SNPs can identify more than is significant (FP)
▪️ Advances in power means previously significant variants may not be anymore
▪️ Problems with diagnosis (e.g., pleiotropy)
What is pleiotropy?
One gene causes different phenotypes (e.g., TREM2 in AD)
(Equally one phenotypes may be caused by different combinations of genes)
What are the main benefits of GWAS?
▪️ Very large studies - detect small effect size variants
▪️ Compare populations
▪️ Inform molecular pathways underlying disease
▪️ Focus for treatment?
What does a negative OR suggest about a variant?
It is protective for disease
How do we report the effect size of an associated SNP?
▪️ Odds ratio
▪️ Forest plot
Why might the frequency of alleles vary across populations (ethnicities)?
▪️ Depending on when they arose in evolutionary history
▪️ Maintenance and representation in populations through time
What techniques can we used to identify variants with small effect sizes?
▪️ GWAS
▪️ Sequencing analysis
What techniques can we use to identify rare variants with small effect sizes?
▪️ Sequencing analysis
▪️ Possibly GWAS imputation if large enough sample size
What are the two main methods of sequencing analysis?
Next Generation Sequencing:
▪️ Whole exome sequencing (only the genome for the coding sequence)
▪️ Whole genome sequencing (both coding and non-coding)
What are the main issues with next generation sequencing?
▪️ Enormous multiple testing issue (millions of SNPs)
▪️ Complex data analysis
BUT very good for rare variants with small ES
How can we improve the number of identified risk genes, filling in the “missing” heritability?
▪️ Improve power of studies (larger sample sizes to detect small effects)
▪️ More genetic markers (currently looking at proxy SNPs but we need to find direct association with causative variants)
▪️ Better phenotyping of psychiatric disorders (endophenotypes?)
▪️ Identify other types of common genetic variants (not just SNPs!)
