Discussion Forum - Public Health Genetics

Question 1

Q1. Some key concepts in genetic epidemiology are “penetrance” and “allelic heterogeneity”: explain these terms.

Answer 1

Penetrance: refers to the degree or proportion of phenotypic expression or manifestation of a particular allele in a population. Can be complete (all individuals who possess the particular alleles express them) or incomplete (only a particular proportion of all those who possess the allele(s) will express them).

Allelic heterogeneity: refers to the fact that the phenoype may be caused by different alleles within the same causative gene, and may arise from differnt (ie a spectrum of differnt) mutations in those alleles

Question 2

Q2. GWAS studies are often said to be agnostic. Explain what this means.

Answer 2

They have been conducted without firm prior hypotheses about a given gene or genes that may underly the disease. The chip based technology permits the prevalence of e.g ~500,000 SNPs to be compared between cases and controls IN ONE GO. This may, on one sense be a strength for discovery sciences but brings with it the difficulty in adequately adjusting the threshold for statistical signicance (notionally the p value should be ~ [0.05 / 500,000] approximately, in this example, so the threshold for GWAS is 10^-8 as you are completing many hypothesis tests in one go.)

Question 3

Q3. For any statistically significant positive association found in a GWAS study there are broad categories of explanation. What are they?

Answer 3

Chance – due to a badly designed study (experimental and/or measurement error comes into play) or artefactual results
Bias - the observed association be due to bias in the manner in which the subjects were selected and followed up or in the way information was obtained from them
Confounding – e.g. stratification; differences between the groups in the distribution of another variable (confounder) that was not measured or taken into account in the analyses could cause a significant positive association
Causal – true causal variant on allele/marker

(i) chance (ii) bias; (iii) population stratification; (iv) Linkage Disequilibrium between the SNP and a Causal Variant; (v) the identified SNP is the Causal Variant

Question 4

Q4. Some findings in GWAS studies are later labelled as part of the “incidentalome”. Explain what is meant by this.

Answer 4

Incidentalomes refer to accidental gene findings which may be of little relevance clinically.
The unexpected findings - SNPs discovered to be linked with certain phenotypes but whose causal and clinical provenance is unknown - you dont know whether they represent a sweetspot for any intervention or whether you can change the outcome.

Question 5

Q5. Why is nature versus nurture the wrong question?

Answer 5

Wrong question for chronic diseases:
Their causes are inherently multifactorial
Gene environment interaction is probably the norm
Genes explain a relatively small fraction of observed variation in risk
Knowing that / what genes are implicated may not change behaviour

Question 6

Q6. We should give our patients their individual DNA profile and their risk of chronic disease when they come to our clinics. Say whether you agree or disagree with this proposition and justify your answer.

Answer 6

I disagree for many of the same reasons:

(i) these profiles are too inaccurate
(ii) ethical problems in obtaining proper “informed consent” and in sharing the data
(iii) negative consequences for patient from insurance companies
(iv) even accurate data may not change risk in predictable ways.