OBQ common mistakes
OBQ · Understand the traps and how to avoid them.
When asked to predict the effect of a mutation, the candidate describes the mutant protein's behaviour but never states what the wild-type protein normally does or why the proposed change follows from that mechanism.
Under time pressure, candidates jump straight to the mutant and treat the normal function as obvious, so the reasoning chain is never built and the answer becomes a guess dressed up as a mechanism.
What to do instead
Before writing the prediction, force out one sentence naming the wild-type molecular function (for example, 'this kinase phosphorylates target X, which releases a transcription factor'). Then state the consequence of losing or gaining that specific activity. If the wild-type sentence cannot be written, the prediction is not yet justified.
The candidate treats transcription and translation as a single step and predicts that a missense or silent change in the coding sequence will alter mRNA abundance or splicing.
Textbook diagrams collapse the central dogma into one arrow, so candidates lose track of which layer a given mutation acts on and apply coding-sequence logic to regulatory questions.
What to do instead
Label every variant by its layer before answering: promoter/enhancer (transcription), splice site (processing), coding (protein sequence), 3' UTR (stability or translation). Ask explicitly 'at which layer does this variant act?' and confine the prediction to that layer unless the question asks for downstream effects.
The candidate confuses genotype with phenotype when interpreting a pedigree or cross, assigning a phenotype directly to an allele symbol without checking dominance, penetrance, or sex linkage.
Allele notation is learned as shorthand, and candidates stop translating between the symbol and the observed trait, especially when a pedigree shows an unaffected carrier.
What to do instead
For every individual in a pedigree, write genotype and phenotype in two separate columns before drawing any conclusion. Check the proposed mode against every affected and unaffected person, including carriers, and explicitly test whether an unaffected parent of an affected child is compatible with the mode claimed.
When a question gives allele or haplotype frequencies, the candidate multiplies frequencies for linked loci as if they were independent, or applies Hardy-Weinberg proportions to a locus under selection or non-random mating.
The independence rule and Hardy-Weinberg are the most rehearsed formulas, so they are applied by default without checking the assumptions that license them.
What to do instead
Before any frequency calculation, list the assumptions in play (random mating, no selection, no linkage, large population) and check each against the stem. If loci are linked, use haplotype frequencies and recombination fraction rather than the product of allele frequencies.
The candidate answers an enzyme-kinetics or metabolic-flux question by reciting Km and Vmax definitions instead of using the values to reason about saturation, rate-limiting steps, or the direction a reaction will run.
Definitions are easy to recall and feel like content, so candidates produce them as a substitute for the quantitative reasoning the question actually tests.
What to do instead
Convert the given values into a comparison before answering: what fraction of Vmax is achieved at the stated substrate concentration, and is the enzyme near saturation? For pathways, identify the step with the lowest capacity or largest negative free-energy change and reason from that step. State the comparison explicitly in the answer.
The candidate describes a metabolic pathway as a list of intermediates without stating where energy or reducing equivalents enter and leave, so ATP and NADH yields are misattributed to the wrong compartment or step.
Pathways are memorised as sequences of names, and the accounting of cofactors is treated as a separate fact set rather than part of the pathway's logic.
What to do instead
For each pathway, draw a two-column ledger as you recall it: carbon intermediates on one side, ATP/NADH/FADH2/NADPH on the other, with the compartment noted. Answer yield questions from the ledger rather than from memory of a total number.
The candidate explains a physiological or behavioural trait by naming the organ system or behaviour and then restating the question, without linking structure to function or function to fitness.
Organismal biology rewards descriptive recall, so candidates stop at labelling and never complete the causal chain from structure through function to survival or reproduction.
What to do instead
Use a three-link template for every organismal answer: structure or trait, the functional consequence, and the fitness consequence in the stated environment. If any link is missing, the answer is incomplete; add it before moving on.
The candidate invokes natural selection as an explanation for any biological pattern, including traits fixed by drift in small populations or constrained by phylogenetic history, without testing the alternatives.
Selection is the most familiar evolutionary mechanism, so it becomes a default explanation and competing mechanisms are never considered.
What to do instead
For each evolutionary explanation, run a short checklist: is the trait heritable, is there variation, is there a consistent fitness difference, and is the population large enough for drift to be negligible? If any condition fails, name the alternative mechanism (drift, gene flow, constraint, byproduct) and justify it.
The candidate reads a graph, table, or dataset and reports the direction of a trend without checking sample size, error bars, units, or whether the comparison is within or between groups.
Candidates scan for the headline pattern and skip the metadata that determines whether the pattern is interpretable, especially when the stem implies a conclusion.
What to do instead
Before interpreting, annotate the figure: axes and units, n per group, what the error bars represent, and the comparison being made. State whether the observed difference exceeds the reported uncertainty before drawing any conclusion, and flag when it does not.
When asked to design or critique an experiment, the candidate lists techniques and reagents but omits the control condition, the variable being manipulated, and how the result would distinguish the hypothesis from its alternative.
Technique recall is trained heavily, so candidates produce a methods inventory and assume the logic of the experiment is implied rather than stated.
What to do instead
Structure every design answer as four labelled parts: hypothesis, independent and dependent variables, the specific control that rules out the leading alternative, and the predicted outcome under each hypothesis. If the outcome would be identical under both hypotheses, the design is not yet discriminating and must be revised.