Design a defensible concentration range
Use logarithmic spacing, include controls and place observed responses on both sides of the expected midpoint whenever possible.
PHARMACOLOGY · ASSAY DEVELOPMENT
Dose-response analysis hub
Move from plate quality and concentration design to an interpretable IC50 or EC50 curve without hiding the model assumptions.
For assay scientists, pharmacologists and early discovery teams.FROM INPUT TO INTERPRETATION
A four-step, reviewable workflow.
The sequence is a practical guide, not a substitute for a study-specific protocol.
Serial dilution calculator →
Check assay separation before fitting
Review control variability and plate quality. A smooth fitted curve cannot rescue weak positive and negative control separation.
Fit and inspect the 4PL curve
Keep the raw observations, fitted line, top, bottom, slope and midpoint visible. Treat an extrapolated midpoint as a warning, not a discovery.
Separate potency from interaction
For combination experiments, compare the observed combination response against an explicitly chosen independence or additivity model.
INTERPRETATION GUARDRAILS
What matters beyond the formula.
Bracket the midpoint
A curve is most informative when measured concentrations cover both asymptotes and the transition region.
Inspect residual structure
Systematic departures from the curve can indicate a poor model, preparation error or biology not represented by a 4PL fit.
Report assay context
Cell line, endpoint, exposure time, normalization and fitting constraints are part of the result, not optional decoration.
Common interpretation questions.
Can I report an IC50 if the curve never reaches 50% response?
A fitted value may still be returned, but it is extrapolated and usually weakly supported. Report the limitation and consider expanding the concentration range.
Should I force the top and bottom to 100 and 0?
Only when the assay definition and prespecified analysis justify those constraints. Forced plateaus can materially shift the midpoint and slope.
Is R² enough to validate a dose-response curve?
No. Review data coverage, replicate behavior, parameter plausibility, residuals and whether the midpoint is bracketed.
SELECTED REFERENCES
Start with the method source.
References inform this workflow and do not imply endorsement of Tossora Lab.
Source-informed and internally checked. Independent scientific or regulatory validation is not claimed.
How we review and correct content →