01Give each measurement its own question

Affinity concerns binding to a target. Efficacy concerns the capacity to produce a response after interaction, while potency concerns the concentration or amount associated with a defined effect. Pharmacological efficacy is also distinct from clinical efficacy: a receptor assay is not a treatment-outcome study.Santiago and colleagues — Core concepts of pharmacology education (opens in a new tab)

If a paper reports only one column, leave the others blank. A blank acknowledges a measurement that was not supplied; it is preferable to deriving a biological claim from an unrelated number.

02Read a binding constant with its model

The equilibrium dissociation constant, Kd, is a common affinity measure. Under a simple reversible one-site equilibrium model, the fraction of sites occupied is the free ligand concentration divided by that concentration plus Kd. Lower Kd corresponds to higher affinity under comparable conditions.Santiago and colleagues — Core concepts of pharmacology education (opens in a new tab)Guilding and colleagues — Core concepts of pharmacology: a global initiative (opens in a new tab)

For a hypothetical Kd of 10 nM, a free ligand concentration of 10 nM gives 10 ÷ (10 + 10) = 0.5, or 50% occupancy in that model. At 90 nM the calculation gives 90 ÷ 100 = 90%. These are model calculations, not estimates for any product or biological system.

03A response curve includes the experimental system

A functional EC50 commonly identifies the concentration producing half the maximum response on an agonist curve. Potency can depend on receptor number, coupling and ligand availability as well as the ligand's interaction with the target. It is therefore not simply another name for affinity.Santiago and colleagues — Core concepts of pharmacology education (opens in a new tab)

Imagine two hypothetical cell preparations with the same receptor sequence. A measured response could differ even if a separate binding experiment gave similar affinity estimates. Before treating the result as a contradiction, inspect what each preparation measures and how its output is normalised.

Similarly, a high plateau and a low EC50 are different observations. One describes response size; the other locates a response level along the concentration axis. Neither should be reduced to a single informal score such as strength.

04Do not equate numbers because their units match

Suppose a binding table gives Kd = 10 nM and a functional table gives EC50 = 2 nM. The appropriate first response is to examine both experiments. Declaring that one table must be wrong assumes that half occupancy and half response are necessarily the same event.

Keep the exact metric when copying values into a spreadsheet. A column called activity with all numbers expressed in nM conceals the distinction between binding, activation and inhibition. Add columns for endpoint, biological system and source so the comparison can be reconstructed.

Also retain whether a value is measured, fitted, estimated or reported as a limit. A change in notation should not silently turn a model-derived parameter into a direct observation.

05Build the conclusion from the measurement

• ✓For a binding claim, identify the target, parameter and binding model.
• ✓For a response claim, identify the endpoint, baseline and reference maximum.
• ✓For a comparison, check that the quantities and assay contexts are compatible.
• ✓For a clinical claim, look for clinical evidence rather than a receptor parameter.

The strongest summary is often a pair of sentences: one for binding and one for function. This keeps useful detail visible and prevents a favourable number from acquiring a meaning that its experiment never tested.