Two curves, two different winners
Imagine looking at two curves in a paper. The first starts climbing well to the left of the second, but the second keeps climbing after the first has flattened out. A colleague points at the figure and asks which peptide gave the stronger response. Where would you point?
There are two reasonable places to look. One is the part of the curve that tells you how much material the assay needed to reach a particular response. The other is the plateau, which tells you what level of response the assay reached. The language gets slippery when a discussion uses “stronger” for both.
Before reading the table of fitted values, read the graph’s labels. Is the horizontal axis a concentration, and is it shown on a logarithmic scale? Is the vertical axis a raw signal or a percentage of a reference? Those details tell you what a number in the table can actually describe.
Read the midpoint and the height separately
The relative EC50 is the concentration corresponding to halfway between a curve’s fitted bottom and top. It is a concentration value, so look to the horizontal axis to read it. Emax is the maximum response in the model or assay comparison. It belongs on the vertical axis.
Some tables express the maximum relative to a reference agonist. That can be handy if you know which reference was used and what its value represents. If both curves are instead scaled to their own maxima, both may reach 100% on the page even when their original signals were quite different.
When a paper gives a “50%” value, check what the percentage is relative to. Half of a compound’s own fitted response range and half of a shared reference range can land at different points. An apparently precise label does not settle which convention the authors used.
Work it out with a pencil
Take two invented curves with zero baselines. Curve A levels off at 100 response units and crosses its own halfway point at 20 nM. Curve B levels off at 40 units and crosses its halfway point at 2 nM. For A, halfway is 50 response units. For B, it is 20.
Under the hypothetical assay conditions, B has the lower relative EC50 and A has the higher maximum. Notice that B never reaches 50 units at all. Its midpoint still exists because the midpoint is relative to its own response range.
You can reproduce the reasoning on a published plot without doing the curve fit yourself. Find the bottom and the top, locate halfway on the response axis, and then read across and down to the concentration axis. The exercise shows you what the reported number means. Use the paper’s reported fit and uncertainty for the actual estimate.
Look for the data underneath the fitted line
A common fit describes the curve through four parameters. The lower and upper plateaus describe the ends, the midpoint locates the transition, and the slope describes its steepness. The smooth line is the model’s account of the points the authors measured. If the observed points end while the curve is still rising, the upper plateau can depend heavily on what the fit assumes.
Now look at just the measured points. Do you see enough of the low and high ends to tell where they level off? Is the estimate inside the tested range? The Assay Guidance Manual discusses that last distinction in its IC50 reporting guidance. It is a useful question to carry over to an agonist curve.
If two estimates are close, extra digits make them look easier to compare than they are. Check whether the paper gives uncertainty and what the error bars summarize. A graph with one point per concentration and a clean line provides less information about repeatability than a figure that explains how the estimates varied across experiments.
Sources: Beck and colleagues, Assay Guidance Manual, NCBI Bookshelf (2017): Assay Operations for SAR Support
See the distinction in a tirzepatide figure
Willard and colleagues ran competition-binding experiments and cAMP response assays using engineered cells expressing human receptors. At the GLP-1 receptor in the albumin-free comparison, they reported roughly fivefold lower competition-binding affinity for tirzepatide relative to native GLP-1, alongside roughly twentyfold lower cAMP potency in the low-receptor-density assay.
The two values appear close together in the paper, but the instruments tracked different things. One comparison concerns binding competition. The other concerns how much cAMP the assay recorded as a response. Figure 1 and Supplemental Table 1 tell you which experiment produced each value.
The paper discloses Eli Lilly employment and possible stock ownership among several authors, and research support and advisory or committee roles among others. The figures describe the specified cell assays and study materials.
Put the assay name beside the number
If you take one note from a concentration-response paper, write it as a sentence about a comparison. “In this receptor assay, against this reference, the authors reported this midpoint on this response scale.” Then put the reported maximum in the next sentence. That gives someone else enough context to see how both statements can be true.
Leave an empty space for uncertainty when the paper does not provide it. A blank is more informative than quietly carrying a best-fit number into a new assay as though it were a property of the material in every setting.
Frequently asked questions
Can a peptide have a lower EC50 and a lower Emax?
Yes. It can reach half of its own response range at a lower concentration while leveling off at a lower maximum. The two numbers refer to different axes.
What if the points never reach a plateau?
Check whether the reported top was fitted or fixed and whether the paper shows uncertainty. The plotted line may extend beyond the observations, so the estimate needs context before you compare it with a well-sampled curve.
Sources: Beck and colleagues, Assay Guidance Manual, NCBI Bookshelf (2017): Assay Operations for SAR Support
Can I compare EC50 values across two papers?
Start by comparing the endpoint, the assay system, the reference and the fitting convention. If those differ, each value is best read within its own experiment. A difference in the table could reflect the measurement context as well as the compound.
Sources: Beck and colleagues, Assay Guidance Manual, NCBI Bookshelf (2017): Assay Operations for SAR Support · Willard and colleagues, JCI Insight (2020): Tirzepatide is an imbalanced and biased dual GIP and GLP-1 receptor agonist
Sources
- Glossary of Quantitative Biology TermsDevanarayan and colleagues, Assay Guidance Manual, NCBI Bookshelf · 2014
- Assay Operations for SAR SupportBeck and colleagues, Assay Guidance Manual, NCBI Bookshelf · 2017
- Tirzepatide is an imbalanced and biased dual GIP and GLP-1 receptor agonistWillard and colleagues, JCI Insight · 2020 · DOI: 10.1172/jci.insight.140532
Published by EliteGen Research for research education. Studies describe their own materials and experimental conditions; they do not establish the identity, quality, or suitability of an EliteGen product. This article does not provide guidance for personal, human, or veterinary use.



