What to expect from a molecular biology viva
There is no single international molecular biology viva format. A viva may assess a research project, a taught course, a laboratory practical or a postgraduate application, and the structure depends on the institution and assessment. Check your course handbook or the official assessment guidance for the format, duration and criteria that apply to you; do not assume that another university’s format is yours.
The questions below are for practising the oral skills common to molecular biology assessments: explaining a method, interpreting results, identifying controls and limitations, and proposing a defensible next experiment. They are not a claim about a particular examination’s questions or marking scheme.
The MySummaries examiner strip can be configured for a molecular biology practice viva like this:
Treat those emphases as practice feedback, not official examination criteria. A useful answer is not a recital of every fact you know. It gives the examiner a clear line of reasoning, then supports that reasoning with the relevant molecular detail.
A repeatable method for answering aloud
Use this sequence when you hear a question:
- Answer the question first. Give your conclusion or the main principle in one sentence.
- Explain the mechanism or method. Name the key components and what they do.
- Interpret the evidence. State what the result supports and what it cannot establish.
- Name controls and limitations. Choose controls that test the specific source of error at issue.
- Suggest the next step. Make it an experiment that would distinguish between plausible explanations.
For example, if asked whether a lower qPCR signal proves that a gene is downregulated, do not answer “yes” from the signal alone. Explain how you would normalise the target to a suitable reference gene, compare the normalised values between conditions, and check that the assay and controls are valid. Then distinguish a change in transcript abundance from a change in protein abundance or biological function.
Keep the first answer compact. If the examiner asks a follow-up, add the detail that addresses it rather than restarting your explanation. If you are uncertain, state what you do know, identify the uncertainty and describe how you would test it. Guessing a precise mechanism or treating one result as proof can be less convincing than a careful, bounded interpretation.
A board for these practice stations can bring the relevant methods and interpretation points together:
- ΔCt = Ct(target) − Ct(reference)
- ΔΔCt** = ΔCt(treated) − ΔCt(control)
- Relative expression = 2^−ΔΔCt if amplification efficiencies are suitable and comparable
- Check protein concentration and loading before comparing bands
- A loading control should be in the linear detection range
- An absent band can reflect extraction, transfer or antibody failure, not only absent protein
- No-template control checks reagent contamination or primer-dimer signal
- No-reverse-transcriptase control checks for genomic DNA contribution
- Validate reference-gene stability across the conditions being compared
State competing explanations first; choose a control or orthogonal assay that separates them. A rescue experiment can test whether a phenotype is linked to the edited gene.
- SpCas9 commonly recognises an NGG PAM adjacent to its guide target
- Target an early constitutive coding exon; confirm edits by sequencing
- Use independent guides and assess predicted off-target sites
Use the board as a prompt for spoken explanations, not as a script to read. For each section, practise explaining why the detail matters and what conclusion it permits.
Station 1: interpreting an RT-qPCR result
Start by identifying the comparison and the normalisation step. In the example, the treated sample has a higher target Ct while the reference-gene Ct is unchanged. That pattern is consistent with less target transcript, but only if the assay is valid and the reference is stable. A strong answer also distinguishes the calculation from the biological conclusion: a fold-change estimate does not by itself establish a functional effect.
A full practice attempt, marked against platform feedback criteria rather than an official examination rubric, looks like this:
Examiner
In control cells, the target Ct is 25 and the reference-gene Ct is 20. In treated cells, the target Ct is 28 and the reference-gene Ct is 20. Assuming comparable amplification efficiencies, interpret the result and tell me what you would check before concluding that treatment reduced target expression.
Correctly normalised the Ct values and interpreted the direction of change.
ImproveState the ΔCt and ΔΔCt values explicitly, then give the fold change.
Mentioned the no-template control and replicate measurements.
ImproveAdd a no-reverse-transcriptase control and check that the reference gene is stable across conditions.
Avoided claiming that the result proves a change in protein function.
ImproveMake clear that transcript abundance is not a direct measurement of protein abundance or phenotype.
A strong answerThe control ΔCt is 25 − 20 = 5, and the treated ΔCt is 28 − 20 = 8. The ΔΔCt is therefore 8 − 5 = 3, giving a relative expression estimate of 2^−3, or 0.125, in treated versus control cells. That is consistent with an approximately eightfold lower target transcript level, assuming comparable amplification efficiencies and valid normalisation. I would check biological replicates, reference-gene stability, a no-template control and a no-reverse-transcriptase control. I would then measure protein or a relevant phenotype if the question concerns functional consequences.
That is a MySummaries station, filled with molecular biology material. Yours is written from your own notes. Start free
The important arithmetic is straightforward: control ΔCt is 25 − 20 = 5; treated ΔCt is 28 − 20 = 8; ΔΔCt is 3; and 2^−3 = 0.125. Avoid saying “the gene is eight times less active”: the calculation estimates relative transcript abundance under its assumptions. The strongest distractor in this kind of discussion is often to treat a Ct difference as a direct fold-change without first normalising it.
If the examiner asks what could undermine the comparison, be specific. Genomic DNA can contribute signal, so a no-reverse-transcriptase control is useful. Contamination or primer-dimer signal can affect a no-template control. A reference gene that changes with treatment can make normalisation misleading. Technical replicates help assess measurement consistency; independent biological replicates are needed to assess variability between biological samples.
Station 2: designing a CRISPR-Cas9 knockout
A good design answer follows the whole path from target choice to evidence of a knockout. It should not stop at “use a guide RNA”. Explain how you would select and validate the edit, and how you would distinguish an on-target effect from an off-target or clonal effect.
A practice station on guide choice and validation can be marked like this:
Examiner
You want to knock out a protein-coding gene in cultured cells using SpCas9. How would you choose a guide, generate and validate edited cells, and test whether a resulting phenotype is caused by loss of this gene?
Proposed an early coding exon and checked for a compatible PAM.
ImproveMention that the exon should be constitutive across relevant transcripts and compare more than one guide.
Included sequencing of the target region.
ImproveExplain how you would assess indels and protein loss rather than relying on a predicted edit alone.
Recognised that a phenotype needs attribution to the intended edit.
ImproveAdd a rescue with an edit-resistant gene construct and consider independent edited clones.
A strong answerI would identify an early coding exon shared by the relevant transcripts and select a guide adjacent to an SpCas9-compatible NGG PAM. I would check predicted off-target sites and use independent guides where practical. After editing, I would sequence the target region to characterise indels and assess loss of the protein with an appropriate protein assay; a genotype alone does not guarantee a functional knockout. I would compare independent edited clones with matched controls and check likely off-target sites. To test whether the phenotype is caused by loss of the gene, I would attempt rescue by re-expressing an edit-resistant version and ask whether the phenotype is restored.
When explaining guide design, say that SpCas9 commonly requires an NGG PAM adjacent to its target sequence. Do not imply that a predicted high-scoring guide is guaranteed to work. The target should be chosen with transcript structure in mind: an early coding exon shared by the relevant isoforms is often a practical choice for disrupting protein production, but the gene’s biology may make another strategy more suitable.
For validation, separate three questions: did editing occur, what edit was made, and was protein function lost? Amplicon sequencing can characterise edits at the target region; an appropriate protein assay can check whether protein is reduced or absent. A phenotype seen in one clone is not enough to establish causation. Independent guides or clones, matched controls and rescue provide stronger evidence. If asked about off-target effects, say how you would assess plausible sites rather than claiming that computational prediction eliminates the risk.
Station 3: troubleshooting a missing western blot band
This question tests whether you can avoid jumping from a failed detection signal to a biological conclusion. Work from sample preparation through detection, and use controls that identify where the signal was lost. Do not list every possible cause without prioritising the checks.
A worked oral attempt can show how to structure that troubleshooting:
Examiner
Your cells show target-gene RNA by RT-qPCR, but you cannot detect the corresponding protein band by western blot. How would you investigate the discrepancy?
Recognised that RNA detection does not guarantee detectable protein.
ImproveState that the band absence alone cannot establish that the protein is absent from the cells.
Mentioned protein extraction and antibody performance.
ImproveCheck total protein, transfer and detection controls in a defined order.
Proposed checking another protein assay or antibody.
ImproveUse a known positive sample and verify the expected molecular-weight region before repeating the experiment.
A strong answerI would not infer that the protein is absent from the cells from a missing band alone. First I would check that the sample contains measurable protein and that the target is present in a known positive control. I would verify transfer and detection with an appropriate total-protein or loading check, then confirm that the primary and secondary antibodies are suitable and that the expected molecular-weight region was examined. I would consider whether the protein is low abundance, unstable or expressed under different conditions from the RNA measurement. If necessary, I would repeat with an orthogonal protein-detection method and compare matched samples collected under the same conditions.
The order helps make the answer efficient. First check whether protein was extracted and loaded; then whether it transferred and was detected; then whether the biological explanation is plausible. An RNA measurement and a protein measurement may differ because transcription and protein abundance are not identical, and because the samples, timing or assay sensitivities may differ. A known positive sample can test whether the antibody and detection workflow can produce the expected signal.
The phrase “I would optimise the blot” is too vague on its own. Name the failure point you would test and the evidence you would look for. If transfer is uncertain, inspect a transfer or total-protein check. If antibody performance is uncertain, use a suitable positive control and verify the expected band region. If the technical checks work, then discuss biological explanations and whether an orthogonal assay would help.
Review your answer, not just your score
After each practice attempt, note one point you explained well, one detail you missed and one follow-up question that would test your reasoning. Use the transcript to locate vague phrases: “check the controls”, “confirm the result” and “do another assay” all need a specific next step. Replace them with a named control or assay and say what result you expect under competing explanations.
A transcript can make one lost mark easier to diagnose:
I would start by checking the target and reference Ct values and calculating the normalised change. Then I would check the controls. I would also measure the protein if the biological question is about function. The result proves the treatment has stopped the gene working.
Then I would check the controls.
The answer does not say which controls address contamination, genomic DNA or reference-gene instability.
Say: Name the no-template and no-reverse-transcriptase controls, and check that the reference gene is stable between conditions.The aim is not to memorise a polished paragraph. Practise adapting the structure to a new question: identify the claim, say what evidence supports it, name the uncertainty and choose the next experiment to reduce that uncertainty. For project vivas, also practise defending your own choices: why this method, why these controls, what alternative you considered and what you would change if the result were different.
How MySummaries helps
MySummaries turns your molecular biology notes into a revision board, then uses that material to create viva stations and recorded practice with examiner-style feedback. You can use your own lecture slides, PDFs or photographed notes to build questions around the methods and concepts you are expected to discuss, then revisit the points your answers missed.