Complex Disease Flashcards
Is there a simple mendelian vs complex disease divide?
No, even Mendelian disease have variable penetrance. As the number of genes affecting a phenotype increases the variability of that genotype and phenotype also increases.
What is the threshold model?
The more disease risk allele you have the higher chance you have of getting that disease. Once you are over the threshold you get that disease. Environmental factors can move where that threshold lies.
How can we distinguish the contribution environment and genetics make to a phenotype?
Heritability is calculated using twin studies. It measures the proportion of a phenotype that an be attributed to variation in genetics. There is both broad and narrow sense heritability.
However some monozygotic twins differ in shared amniotic environment.
Why are adoption studies useful?
Compare adopted children to natural parents. This allows comparison of disease rates when the environment is completely different but genetics are similar and so calculate the environmental contribution to disease. Can also do migrant studies in a similar way
Issues: prenatal environment, age at adoption and the matching of parents to children.
But easier to do than twins as more common.
How do we determine the genetic architecture of a disease?
Segregation analysis in families to determine the most likely scenario. This includes inheritance but also whether it involves rare high effect or common low effect etc. Example being Hirshprung disease (chronic constipation due to nerve problem) was predicted to be dominant and later confirmed to be.
How can we find genes that might be influencing a disease?
Linkage analysis - there are 2 types, parametric and non parametric (affected sibling pairs).
Generally these involve lots of assumptions but not too many people. As a result it is good for high penetrance diseases.
Association studies - case controls and GWAS. These work well with common variation. Very few assumptions made but lots of people required.
Describe the characteristics of parametric linkage studies and when is it useful in complex disorders?
Parameters required: mode of inheritance, gene frequency, penetrance, phenocopy rate and a RELIABLE DIAGNOSIS. Note that alot of this might be found in segregation analysis.
Strengths: only need small extended family, does not need dense markers, able to exclude genomic regions (association studies can’t do this).
Weaknesses prior information and sensitive to miss diagnosis.
In complex disorder it can be useful to identify mendelian subtypes e.g. Alzheimer’s helped confirm Abeta by finding APP and PSS that may help to explain sporadic cases. IN Breast cancer found BRCA genes but not that useful in sporadic case but did highlught important pathway. Can also assess isloated populations or disease with a good endophenotype e.g. hypercholesterolaemia and CAD and the Amish and genes for healthy ageing.
Describe the characteristics of non parametric linkage studies and… when is it useful in complex disorders?
This is done using affected sibling pairs.
Strenghts: less parameters and sibling pairs are common
Weaknesses: maps to large loci, difficult for late onset such as AD.
Definition: individual concordant for a given genetic trate should show greater than expected concordance for alleles of a marker to which it is linked i.e. there should be more allele IBD than you would expect. This does mean you need to genotpye parents.
This was used successfully in psoriasis to identify HLA-C as a recessive allele.
What are association studies?
These studies suggest that if allele X is associated with a high frequency in disease individuals than controls then it is likely to be influencing that disease and consequently having that allele increases the risk of having that disease.
What can cause an association to an an allele?
Allele X causes the disease
Allele Y protects from the disease
Populations stratification e.g. chopstick allele
Allele X is in LD with the causative allele
Type 1 error.
What are the strengths and weaknesses of association studies?
No need to specify disease model, case control samples are easy to acquire, and you can get high genetic resolution.
Weaknesses: do you chose candidate genes or type SNPs and can’t exclude genomic regions.
Compare Linkage analysis with association studies
Linkage: Disease linked to loci linkage within families Locus heterogeneity is problem Finds strong rare effects
Association: Disease associated with alleles Association within a population Allelic heterogeneity is a problem Finds weak common effects.
What is the common disease common variant hypothesis - what is the alternative and why is this a problem?
This is what association studies assume
Alleles that cause disease will be common in the population and so will also be common in the non diseased. It does not mean that all cases necessarily have the same variation. It simply means that diseased individuals have lots of different small risk alleles.
This implies that causative allele are old and have spread widely - selection?
Could be complex disease is caused by lots of very rare variants. Association studies may not find these alleles because there isn’t enough people with that allele or because it isn’t well tagged by SNPs.
How do association studies actually work?
Usually a SNP isn’t the actual causative variant. Associations studies works becasue the SNPs tested are in LD with the causative variant.
Two loci are in LD if the frequency of haplotypes vary from that expected by their allele frequencies. LD decays by recombination on a large scale this is linear to distance but on a small scale step wise due to recombination hot spots are in the genome.
What types of association studies are there and what kinds of scale can they work on?
Case control vs quantitative trait and candidate gene/locus vs GWAS.
Scale of studies:
Candidate polymorphisms - singe SNP
Candidate genes - multiple SNPs in candidate genes
Fine mapping - high resolution scan across linkage interval
Genome wide - SNPs across whole genome. These SNPs can be chosen randomly or coding SNPs only.
What must be considered when conducting a GWAS study?
Sample size needed Number of SNPs to type Are these SNPs in HWE - if not get rid Imputation How are you going to analyse?
