Reading concentration–response curves

The horizontal axis may represent concentration on a linear or logarithmic scale. The vertical axis may show raw signal, change relative to a control or a normalized percentage. Before comparing curves, check that they express the same variable and that normalization uses the same reference.

An apparently shifted curve may reflect a different scale. It also matters whether the graph shows replicates, averages or only the fitted line: each format reveals a different amount of experimental information.

The 2006 quantitative screening research compared concentration series with a single-concentration approach. In an enzyme assay, the authors found activities that the single-point format could miss.

A curve is useful for more than obtaining a central estimate. It can show whether there is a reproducible transition, whether a plateau is approached and whether behavior appears only at one extreme. These features distinguish patterns that an active-or-inactive checkbox improperly combines.

An ecotoxicological assay study derived a simplified logistic equation from curves with similar slopes. The simplification was supported by that experimental dataset, not by a universal rule applicable to every assay.

Fixing a slope or plateau can facilitate fitting, but it also constrains the result. Identify which parameters were estimated and which were imposed when reading a publication. An incomplete experimental range does not become complete because software returns a curve.

Research on luciferase assays demonstrated a counterintuitive phenomenon. In cellular models, certain inhibitors stabilized the reporter protein and produced an apparent activation signal. The response could belong to the detector rather than the intended biological process.

A curve's shape therefore does not demonstrate specificity. Confirmation using a different readout helps separate activity on the system of interest from interaction with the measurement tool. The relevant control depends on the reporter used.

Interpretation must follow the range actually observed. Without a sufficient plateau, the maximum and midpoint may be poorly determined. If variability is substantial, small differences between curves may not support confident ranking of the materials.

A useful description states shape, range covered, variability and uncertain parameters. That information allows the fit to be assessed and later experiments compared without presenting every number produced by software as a fixed property of the molecule.