1. Start with the two fitted midpoints
The worked plate contains a reference standard and a test sample, each measured at eight dilution levels with three replicate wells. Both series use concentration in ng/mL and the same increasing common-shape 4PL convention.
Keep the assigned reference potency separate from plate concentration. In this example it is 100%, so the final ratio is already expressed as test potency relative to the reference.
| Input | Worked value | Meaning |
|---|---|---|
| Reference EC50 | 20.000 ng/mL | Reference midpoint |
| Test EC50 | 25.000 ng/mL | Test midpoint |
| Assigned reference potency | 100% | Multiplier for the reported scale |
| Curve convention | Increasing common-shape 4PL | Reference and test use the same comparison model |
2. Check whether one ratio can describe both curves
Relative potency asks how much reference and test material is needed to produce the same response. Read concentration from left to right and response from bottom to top in the plot below. Because concentration uses a logarithmic scale, the same horizontal distance represents the same dose ratio anywhere on the axis.
The worked reference and test series rise through the same response range with the same common-shape 4PL form. The test curve sits to the right: at the shared midpoint response of 987.5, the reference needs 20.000 ng/mL and the test needs 25.000 ng/mL. Here, comparable behavior means that this horizontal dose separation remains consistent enough across the response range for one potency ratio to summarize the relationship.
Do not decide comparability by eye. The common-shape calculation shares shape parameters by design; it does not prove the assumption. Apply the predefined model-specific adequacy tests. If a required check fails, the arithmetic can still produce 80.0%, but the result is not reportable under that method.
The test needs more concentration to reach the same response
- Horizontal axis
- Concentration increases from left to right on a logarithmic scale, so equal distances represent equal dose ratios.
- Vertical axis
- Response increases from bottom to top. EC50 is the concentration at the response halfway between the fitted lower and upper plateaus.
- Observed data
- The circles and squares come from the downloadable fixture. Its three synthetic replicates coincide exactly at every concentration.
- What to compare
- The complete reference and test series should follow similar shapes, not merely cross one selected response level.
View the exact values plotted
| Concentration (ng/mL) | Reference response | Test response |
|---|---|---|
| 100 | 1668.947809 | 1609.303791 |
| 50 | 1444.070414 | 1346.533080 |
| 25 | 1108.946298 | 987.500000 |
| 12.5 | 736.784121 | 628.466920 |
| 6.25 | 437.237792 | 365.696209 |
| 3.125 | 252.577901 | 214.039859 |
| 1.5625 | 156.789557 | 138.241533 |
| 0.78125 | 111.525387 | 103.076923 |
3. Calculate the relative potency
For this increasing-curve convention, divide the reference midpoint by the test midpoint, then multiply by the assigned reference potency. The test needs more concentration to reach its midpoint, so it is less potent than the reference.
Use the stored, unrounded fitted values in the calculation. The three-decimal midpoint display and one-decimal potency display shown here are reporting formats, not intermediate rounding steps.
20.000 ng/mL ÷ 25.000 ng/mL × 100% = 80.0%Recalculate 20 ÷ 25 × 100
This increasing-curve example uses reference EC50 ÷ test EC50 × assigned reference potency. Change the values to see how each input affects the result.
- Illustrative relative potency
- 80.0%
This arithmetic does not establish reportability. Continue to Step 4 and apply the comparability and suitability rules defined for the assay.
4. Apply suitability before reporting the result
For the downloadable example, the configured reference and test fit-scale R-squared minimums, relative-potency range, independently fitted Hill-slope ratio range, and maximum replicate CV all pass. The report shows every criterion, observed value, configured limit, and PASS result beside the 80.0% calculation.
Those limits belong to this synthetic example. A team must define and justify its own model, comparability rules, acceptance limits, rounding rule, and failure response for the assay’s intended use.
5. Preserve enough evidence to reproduce it
Retain the source CSV and its fingerprint, well assignments, dilution table, individual observations, exclusions and reasons, model settings, fitted parameters, plots, suitability outcomes, unrounded calculation inputs, reported result, software version, and analysis configuration fingerprint.
The downloadable plate and the analysis link below reproduce the same 20.000 ng/mL, 25.000 ng/mL, and 80.0% example. They demonstrate the software workflow; they do not validate or qualify a customer’s configured assay workflow.
Limits and other method conventions
This formula is specific to the increasing-curve convention stated above. Decreasing curves, a different model parameterization, an assigned reference potency other than 100%, or preparation and predilution corrections can change the defined calculation. Put those choices in the controlled method instead of selecting them after seeing a result.
Alternate joint models, shared-parameter fits, parallel-line methods, and other validated procedures may estimate relative potency differently. Whatever method is selected, a failed comparability or suitability rule remains a failure; changing the equation must not rescue the run.
Reproduce the 80.0% result
Download the source CSV, open the same plate in Provenarium, and inspect the retained settings, fitted curves, suitability outcomes, and report produced by the current release. The fixture supports software testing and demonstration; it is not method-validation evidence.
Frequently asked questions
Should rounded EC50 values be used in the ratio?
No. Calculate with the stored fitted values and round only the displayed result according to the predefined reporting rule.
Is 80.0% a universal acceptance limit?
No. It is the calculated result for this synthetic example. Each team defines and justifies its assay-specific acceptance range before use.
Does reproducing this example validate my assay?
No. It demonstrates a traceable software workflow. Your organization must validate or qualify its configured workflow as fit for its intended use.