01A preference needs a comparison

IUPHAR describes receptor selectivity as relative. A ligand can favour one receptor subtype while still affecting others under different concentration conditions. Binding selectivity and functional selectivity across targets should also be distinguished by the measurement used.IUPHAR Pharmacology Education Project — Pharmacodynamics (opens in a new tab)

A claim about receptor A compared with receptor B does not automatically include receptor C. This seems simple, but a summary can easily drop the comparison and turn preferential activity within a two-target experiment into apparent specificity across biology.

02Check what a selectivity ratio divides

Suppose a hypothetical matched binding study reports dissociation constants of 5 nM at receptor A and 100 nM at receptor B. Under that study's binding model, the smaller constant corresponds to higher affinity. The ratio 100 ÷ 5 = 20 expresses a twenty-fold preference by that measure. It is not a twenty-fold difference in every possible biological response.Guilding and colleagues — Core concepts of pharmacology: a global initiative (opens in a new tab)

Write down the numerator and denominator. Reversing them produces 0.05, which describes the same numbers in a different direction. A bare value of 20 without the metric or target order is difficult to audit and can be copied incorrectly.

Do not divide a binding constant from one target by a response concentration from another and call the result the same selectivity measure. Matching the unit nM is insufficient when the quantities themselves differ.

03An untested receptor is an unanswered question

Draw a small matrix when reading a screening paper: targets down the side, measurements across the top. Mark measured, below the stated detection criterion and not tested separately. An empty cell is not a negative result.

For example, an assay panel containing A and B might show a clear preference for A. A later experiment may test C and reveal an interaction there too. The first result has not necessarily become false; the broader claim that A was the only possible target was never supported by that panel.

This distinction also applies to a search that finds no paper. Not finding a reported interaction cannot establish that the interaction was experimentally excluded. Keep search coverage separate from assay coverage.

04Use database summaries to locate the underlying work

The IUPHAR/BPS ligand activity charts separate data by targets, species and activity types, and provide original references. Their guidance recommends checking the source publication. A plotted summary is therefore a route into the evidence, not a reason to ignore the experiment behind each point.IUPHAR/BPS Guide to Pharmacology — Ligand activity chart guidance (opens in a new tab)

Before combining values, check whether they involve the same species, receptor construct and kind of assay. A tidy visual ranking can contain heterogeneous measurements. Preserve those differences in your notes rather than selecting the most favourable number from each entry.

If the original publication is unavailable, say which database record you could inspect and which details remain unverified. A source limitation is more informative than a confident numerical comparison built from incompatible fields.

05Choose wording that matches the panel

• ✓State the preferred receptor and the alternatives actually examined.
• ✓Name the metric and retain its units or logarithmic notation.
• ✓Keep assay conditions and species visible when comparing studies.
• ✓Distinguish not tested from no detected activity under stated conditions.

A bounded statement such as preferential binding to A over B in the reported assay communicates a useful finding. It leaves room for further targets and different experimental contexts without overstating what the study established. It also avoids treating selectivity as a substitute for safety or clinical effectiveness.