POPH90111 Genetic Epidemiology
POPH90111 Overview
- University of Melbourne
- S1 2026
- 7 chapters
- Health and Medicine
Genetic Epidemiology teaches how to measure the genetic contribution to disease and to appraise the study designs that produce the evidence — familial aggregation, twin heritability, genetic association and GWAS, Mendelian randomisation, penetrance, gene–environment interaction and genetic screening.
- Assessed by Online MCQ 10% · Assignment 1 (written, take-home) 40% · Assignment 2 (written, take-home) 50%
- Key terms Familial aggregation, Heritability (h²), Genome-wide significance, Mendelian randomisation
- How to prepare Treat the subject as a pipeline of paired skills — for every module…
- Most asked What does POPH90111 cover?
What you get in the full POPH90111 Exam Bible
- 7 chapters, 52 pages — every examinable topic
- 14 original practice questions, each with a model answer in the Assignment 2 format
- 7 worked examples, solved the way POPH90111 marks
- 35 computed figures, redrawn from the course’s own numbers
Free on this page: assessment table, chapter map, one worked example, key terms, FAQ. The rest is in the full exam bible.
How POPH90111 is assessed
| Component | Weight | Format |
|---|---|---|
| Online MCQ | 10% | Open-book · ~30-min LMS quiz of ~10 questions · Introduction + Module 1 · completable any time within a one-week window |
| Assignment 1 (written, take-home) | 40% | Modules 1–3 · submitted online over a multi-week window |
| Assignment 2 (written, take-home) | 50% | Modules 4–8 · submitted online over a multi-week window — confirm the exact weights and dates in your subject guide |
What other students are asking about POPH90111
- Where can I find lecture notes or study guides for POPH90111 at UniMelb?
- The POPH90111 study guide in AskSia Library covers Genetic Epidemiology in 7 chapters, with the assessment table, a chapter-by-chapter topic map, key terms and a worked example, free to read on this page.
- Is there a final exam in POPH90111?
- No final exam is listed in the offering this AskSia Library guide is built from; POPH90111 is assessed by Online MCQ 10%, Assignment 1 (written, take-home) 40%, Assignment 2 (written, take-home) 50%.
- What does POPH90111 cover on heritability?
- In AskSia Library's POPH90111 guide, the Heritability chapter explains variance partition · twin studies · Falconer · ACE model · liability threshold.
- What does POPH90111 cover on genetic association studies?
- In AskSia Library's POPH90111 guide, the Genetic Association Studies chapter explains case-control SNP · allelic χ² · GWAS · 5×10⁻⁸ · Manhattan & QQ plots.
- What does POPH90111 cover on mendelian randomisation?
- In AskSia Library's POPH90111 guide, the Mendelian Randomisation chapter explains instrumental variables · the three IV assumptions · the Wald ratio · pleiotropy.
What POPH90111 covers
Seven chapters → one assignment-ready map of the genetic-epidemiology pipeline. Each links to its free chapter guide.
Genetic Epidemiology Foundations
Genes vs environment · HWE · LD · pedigrees · familial aggregation (OR/RR/SMR)Genetic epidemiology asks one question first: do genes contribute to who gets a disease?
- 1.1 Genes vs environment — why families resemble each other
- Degrees of relatedness (1st ½, 2nd ¼, 3rd ⅛)
- 1.2 The vocabulary: alleles, genotypes, haplotypes; germline vs somatic
- Modes of inheritance defined on risk (dominant / recessive / codominant)
- 1.3 Hardy–Weinberg equilibrium: p² + 2pq + q² = 1, carrier frequency, HWE as QC
- 1.4 Linkage disequilibrium: D, D′ and r² — the tagging metric
Heritability
Variance partition · twin studies · Falconer · ACE model · liability thresholdFamilial aggregation told us a trait runs in families; heritability asks the sharper, quantitative question: of all the variation in a trait across a population, what proportion is due to genetic differences between people?
- 2.1 Heritability is a proportion of variance (V<sub>P</sub> = V<sub>G</sub> + V<sub>E</sub>; h² = V<sub>A</sub>/V<sub>P</sub> ≤ H²)
- 2.2 SD → variance (V<sub>P</sub> = SD²); estimate separately by sex and zygosity
- 2.3 Why twins? Holding the environment still (MZ ~100% vs DZ ~50% genes)
- 2.4 Falconer’s formula h² = 2(r<sub>MZ</sub> − r<sub>DZ</sub>)
- 2.5 The ACE model: A = 2(r<sub>MZ</sub>−r<sub>DZ</sub>), C = 2r<sub>DZ</sub>−r<sub>MZ</sub>, E = 1−r<sub>MZ</sub>
- 2.6 The liability-threshold model for binary traits
Genetic Association Studies
Case-control SNP · allelic χ² · GWAS · 5×10⁻⁸ · Manhattan & QQ plotsHeritability said genes matter; association asks which ones.
- 3.1 The idea: exposure = a SNP; three reasons for a signal (causal / LD / stratification)
- Candidate-gene vs genome-wide design
- 3.2 The 2×2: per-allele OR = ad/bc and the allelic χ² (df = 1); genotypic df = 2
- 3.3 Multiple testing and the 5×10⁻⁸ genome-wide threshold (Bonferroni)
- 3.4 Reading the Manhattan plot (peaks above ≈7.3 = loci, not causal genes)
- 3.5 The QQ plot & genomic inflation (λ<sub>GC</sub>); tail-lift vs whole-line lift
Mendelian Randomisation
Instrumental variables · the three IV assumptions · the Wald ratio · pleiotropyObservational epidemiology is haunted by two ghosts: confounding (a third factor drives both exposure and outcome) and reverse causation (the disease changes the exposure, not the other way round).
- 4.1 The problem MR solves: confounding & reverse causation
- 4.2 The instrumental-variable idea (G → X → Y, G acts on Y only through X)
- 4.3 The MR DAG and the three IV assumptions (Relevance, Independence, Exclusion)
- 4.4 The Wald (ratio) estimate β<sub>X→Y</sub> = β<sub>G→Y</sub> / β<sub>G→X</sub>; IVW for many SNPs
- 4.5 How MR avoids confounding and reverse causation
- 4.6 Threats to MR: horizontal pleiotropy (#1), weak instruments, stratification, canalisation
Penetrance and Expressivity
Age-specific penetrance · ascertainment bias · the synthetic cohort · modifiersDiscovering that a variant is associated with disease is only half the story.
- 5.1 Complete vs incomplete penetrance; penetrance by genotype-dose (the inheritance signatures)
- 5.2 Phenocopies: carriers can have risk < 1 and non-carriers risk > 0
- 5.3 Age-specific (cumulative) penetrance — survival analysis with censoring (MSH6 ~50% by 70)
- 5.4 Estimating penetrance — the design determines the answer
- 5.5 The penetrance trap: ascertainment bias & the weighted synthetic cohort
- 6.1 Expressivity — whether vs how badly
Gene–Environment Interaction
Additive vs multiplicative scale · the case-only design · precision preventionGenes and environment do not always act in separate lanes.
- 7.1 What G×E is, and the scale-dependence centrepiece (additive vs multiplicative)
- The four course risks: 0.02/0.04 (gene−) and 0.03/0.06 (gene+)
- 7.2 Computing interaction on each scale (no-mult ⇔ equal RRs; no-additive ⇔ equal RDs)
- 7.3 Synergism vs antagonism
- 7.4 Designs for testing G×E (case-control product term vs case-only)
- 7.5 The case-only design — logic, power and its independence-assumption trap
Genetic Screening
Wilson–Jungner · sensitivity/specificity/PPV · ROC/AUC · NNT & NNSScreening is the end of the genetic-epidemiology pipeline: once we understand a genetic role, have discovered the variants and characterised their risk, the practical question is whether testing asymptomatic people does more good than harm.
- 8.1 Disease vs genetic screening; the Wilson–Jungner principles (WHO, 1968)
- The ‘can test’ ≠ ‘should screen’ gap (principles 7–9 are where genetic screens fail)
- 8.2 NNT and NNS — the course’s signature calculation (ARR → NNT → NNS)
- Why targeting beats population screening (NNS dominated by carrier frequency)
- 8.3 Sensitivity, specificity, PPV/NPV and the prevalence-dependence of PPV
- 8.4 The ROC curve & AUC (sensitivity vs 1−specificity; AUC as discrimination)
The whole subject is a calculation + critical-appraisal toolkit: every assessment item is the same chain — calculate a measure, interpret it, then critically appraise the design that produced it.
Because all three assessments are open-book and take-home (an MCQ and two written assignments, with no closed-book exam), no fact is worth a mark on its own — the marks live in the method, the interpretation and the appraisal. This guide teaches each to assignment standard.
POPH90111 cheatsheet for exam revision
AskSia publishes a free POPH90111 cheatsheet for exam revision, an exam cheat sheet that follows the official permitted-materials rule for this assessment and condenses the topics in this guide. Open the POPH90111 cheatsheet.
Worked example · free
Twin heritability by Falconer’s formula — the signature calculation, mark by mark
- +1Check the direction first. rMZ = 0.78 > rDZ = 0.46, so MZ twins resemble each other more than DZ twins — consistent with a genetic contribution. (If rMZ ≤ rDZ, h² would be about 0 and you would stop here.)
- +2(a) Apply Falconer. h² = 2(rMZ − rDZ) = 2(0.78 − 0.46) = 2 × 0.32 = 0.64. About 64% of the variation in female height in this population is additively genetic.
- +1(b) A (additive genetic). A = 2(rMZ − rDZ) = 0.64 — identical to Falconer, because ACE contains it.
- +1(b) C and E. C = 2rDZ − rMZ = 2(0.46) − 0.78 = 0.14 (shared environment); E = 1 − rMZ = 1 − 0.78 = 0.22 (unique environment + error). Check: 0.64 + 0.14 + 0.22 = 1.00 ✓.
- +1(c) The caveat. A higher MZ-than-DZ correlation is evidence for, but not proof of, an inherited genetic aetiology; heritability is a property of a population in an environment, not of an individual, and the estimate leans on the equal-environments assumption.
Key terms
- Familial aggregation
- The clustering of a disease in families more than chance would allow, measured by an odds ratio, relative risk, standardised morbidity ratio or recurrence-risk ratio λR. It is evidence for — never proof of — a genetic role, because relatives also share an environment.
- Heritability (h²)
- The proportion of the phenotypic variance in a population that is due to additive genetic variance, h² = VA/VP. It describes differences between people in one population and environment, not the trait of any individual. Falconer’s twin estimate is h² = 2(rMZ − rDZ).
- Genome-wide significance
- The strict p-value threshold a genetic association must clear in a GWAS, p < 5×10⁻⁸ (about −log10 = 7.3 on a Manhattan plot). It is the Bonferroni correction for roughly one million independent common-variant tests, and a genuine hit must also replicate.
- Mendelian randomisation
- An instrumental-variable method that uses a genetic variant as a proxy for a modifiable exposure. Because alleles are allocated randomly at conception, the variant is largely free of confounding and reverse causation, so it behaves like the randomisation arm of a trial — subject to its three IV assumptions (relevance, independence, exclusion restriction).
- Penetrance
- The probability that a person with a given genotype develops the disease by a specified age, Pr(disease | genotype, age) — a conditional, age-specific risk. It is usually incomplete (below 1), and clinic-recruited carrier families over-estimate it (ascertainment bias).
POPH90111 FAQ
Is POPH90111 hard?
It is conceptually rich but calculation-light: most marks come from applying a standard measure (an odds ratio, Falconer’s h², a Wald ratio, NNT/NNS) and then interpreting and critically appraising the study design. Because every assessment is open-book and take-home, the difficulty is not memorisation — it is reasoning precisely about bias, scale and what a number does and does not mean.
How is POPH90111 assessed?
There is no closed-book exam. Your grade is three take-home pieces: an open-book online MCQ (about 10%, covering the Introduction and Module 1), Assignment 1 (about 40%, Modules 1–3) and Assignment 2 (about 50%, Modules 4–8). Confirm this year’s exact weights and dates in your subject guide.
What does POPH90111 cover?
The genetic-epidemiology pipeline end to end: foundations and familial aggregation (Hardy–Weinberg, linkage disequilibrium, pedigrees, OR/RR/SMR), twin heritability (Falconer and the ACE model), genetic association and GWAS (allelic χ², the 5×10⁻⁸ threshold, Manhattan and QQ plots), Mendelian randomisation, penetrance and expressivity, gene–environment interaction, and genetic screening (sensitivity/specificity/PPV, ROC/AUC, NNT and NNS).
Do I need a strong statistics or genetics background for POPH90111?
No. The subject builds the genetics vocabulary from scratch and the statistics it needs are arithmetic-level — ratios, a 2×2 table, a variance decomposition, a number-needed-to-treat. The harder skill it trains is critical appraisal: naming the signature bias of each study design and stating the right caveat.
Is using AskSia for POPH90111 cheating?
No. AskSia is a study reference written in our own words — we host none of your lecturer’s files, and Sia teaches you the method to earn the marks; it does not complete or sit your assessments.
Questions students ask in each POPH90111 chapter
- Why does familial aggregation not prove a disease is genetic?
- Because relatives share two things that are hopelessly tangled: their DNA and their environment. Families eat, live and are exposed alike, so a disease that clusters in families could be inherited or simply the product of a shared environment. Aggregation that is stronger in closer relatives and declines with relatedness points toward genes, but shared environment is always the rival explanation — which is why the course wants you to write ‘evidence for, but not proof of, a genetic aetiology’ every time. Chapter 01 · Genetic Epidemiology Foundations
- Why is variance, not the trait value, the thing being explained?
- Heritability is the proportion of the variance — the spread of differences between people — that is genetic. It says nothing about why any one person has the trait value they do; it explains why people differ from each other in one population and one environment. That is why a highly heritable trait (height is ~80% heritable) can still rise with better nutrition, and why h² tells you nothing about an individual. Chapter 02 · Heritability
- Why does a GWAS need a threshold as strict as 5×10⁻⁸?
- Because of multiple testing. Test one SNP at α = 0.05 and you accept a 1-in-20 false-positive risk; test a million SNPs and you expect about 50,000 ‘significant’ results purely by chance. The Bonferroni correction divides 0.05 by the number of effectively-independent tests — about a million common-variant tests across the genome — giving 5×10⁻⁸. Memorise both the p-value and its −log10 ≈ 7.3. Chapter 03 · Genetic Association Studies
- How does a genetic variant escape confounding and reverse causation at once?
- A germline variant is allocated at conception by meiotic chance, so it is essentially independent of the later lifestyle and social factors that confound ordinary studies (the independence assumption) — and it is fixed before any disease exists, so the disease cannot have caused it (no reverse causation). Random assignment of alleles plays the role that random assignment of treatment plays in a randomised controlled trial, which is why MR is called ‘nature’s randomised trial’. Chapter 04 · Mendelian Randomisation
- Why isn’t a carrier of a pathogenic variant certain to get the disease?
- Because penetrance is usually incomplete — below 1 — so some carriers never develop the disease even by old age. Most disease genes behave this way, which is exactly why ‘carrier’ does not mean ‘will be affected’ and why counselling is framed as risk, not certainty. Two boundary facts matter: carriers can have risk below 1 (incomplete penetrance) and non-carriers can have risk above 0 (phenocopies / sporadic cases). Chapter 05 · Penetrance and Expressivity
- How can the same data show interaction on one scale and none on another?
- Because ‘no interaction’ has two definitions. On the multiplicative scale, no interaction means equal relative risks (joint RR = product of the separate RRs); on the additive scale, it means equal risk differences (joint RD = sum of the separate RDs). With the course’s risks (0.02/0.04 and 0.03/0.06), the relative risks are both 2.0 (no multiplicative interaction) but the risk differences are 0.02 and 0.03 (additive interaction present). Same numbers, two answers — which is why you must always state the scale. Chapter 06 · Gene–Environment Interaction
- What does ‘we can test ≠ we should screen’ mean?
- It is the single biggest screening trap. The existence of an accurate test is necessary but not sufficient: a screen only helps if knowing the result reduces disease, disability or death and the benefits outweigh the harms (false positives, anxiety, insurance and legal consequences, variants of unknown significance, overdiagnosis). In the Wilson–Jungner checklist, principles 7–9 — is there an effective intervention, are facilities available, is the cost balanced — are where most candidate genetic screens fail, even when a valid test exists. Chapter 07 · Genetic Screening
How to study for the exam
Treat the subject as a pipeline of paired skills — for every module, learn the signature calculation and the signature critical-appraisal point together, because every assignment item is “calculate → interpret → appraise.” Drill the recurring calculations until they are automatic (build a 2×2 then OR = ad/bc; Falconer h² = 2(rMZ−rDZ) and the ACE solve; the Wald ratio; ARR → NNT → NNS), and rehearse the matching appraisal for each design (recall and selection bias in case-control aggregation, the equal-environments assumption in twin studies, population stratification in GWAS, horizontal pleiotropy in MR, ascertainment bias in penetrance, the independence assumption in case-only G×E, and PPV-at-prevalence in screening).
Because the assessments are open-book, treat the formula boxes and design-appraisal grids as a lookup — the marks are in the interpretation you write, and the one line that recurs everywhere is ‘evidence for, but not proof of, a genetic aetiology.’