NovaSolver›IC50/EC50 Dose-Response Hill Curve Simulator Back
Pharmacology / Drug Discovery
IC50/EC50 Dose-Response Hill Curve Simulator
A real-time Hill-equation simulator for the dose-response curves used everywhere in pharmacology and drug discovery. Sweep IC50, Hill coefficient and the top/bottom response, and instantly read pIC50, therapeutic-index ratio and the response at any test dose.
Drug molecules (blue) bind to the receptor (grey) active site. As the dose exceeds IC50/EC50, occupancy (red) rises and so does the response. Colour indicates response level at the test dose (green = low / red = high).
B = Bottom (baseline), T = Top (maximum response), n = Hill coefficient (1: simple binding, >1: cooperative, <1: anti-cooperative). IC50 is in M (1e-9 = nM). The inhibition branch goes down with dose; the agonist branch goes up.
Concentrations producing 10% / 90% response. A small IC90/IC10 ratio means a steep curve and a narrow therapeutic window.
Pharmacology IC50/EC50 and Hill-equation dose-response curves
🙋
"IC50" appears everywhere in pharmacology papers. What is it actually, and is "smaller = better drug" really true?
🎓
Good question. IC50 is the drug concentration that produces half-maximal inhibition. So yes, smaller IC50 means you need less drug to get to half effect, so the molecule is more potent. A new kinase inhibitor with IC50 = 1 nM is very strong, while IC50 = 10 μM is still a weak hit. EC50 is the agonist counterpart, the concentration that gives half of the maximum response, used a lot for GPCR ligands and endogenous agonists.
🙋
And what about the "Hill coefficient"? When I move the slider I just see the curve get steeper, but what does it really mean?
🎓
The Hill coefficient n tells you the steepness of the curve, and physically the cooperativity of binding. n=1 means drugs bind independently to one site each (classic Michaelis-Menten style). With n=2 or 3 you have positive cooperativity, where the first binding helps the next. The textbook example is haemoglobin binding O2 with n≈2.8. n much less than 1 means negative cooperativity or a heterogeneous population, often a sign of assay noise. Push n from 1 to 3 and you will see IC10 and IC90 collapse towards IC50.
🙋
OK, then what is "pIC50"? Papers sometimes report that instead of IC50 itself.
🎓
pIC50 is IC50 converted to molar and then taking the negative log10, just like pH. IC50 = 100 nM = 1e-7 M gives pIC50 = 7; 1 nM gives pIC50 = 9. Each integer step is a tenfold potency change, which is much easier on the brain. SAR discussions in med-chem groups and patents almost always quote pIC50. Rules of thumb like "pIC50 ≥ 7 is a hit, ≥ 8 is a lead candidate" are very common.
🙋
What about the "therapeutic index IC90/IC10" shown here? Why not just IC90/IC50?
🎓
The classical therapeutic index is TD50/ED50 (toxic 50% over effective 50%), but the IC90/IC10 ratio is a handy proxy for curve sharpness. It tells you how many fold of dose lies between "barely effective" and "almost fully effective". For n=1 the ratio is 81-fold; for n=2 it is 9-fold; for n=3 about 4.3. A small ratio means a sharp all-or-nothing response, which matters for anaesthetics or neuromuscular blockers, where you want a clean switch.
🙋
Finally, why do you bother showing "typical affinity range" for each receptor class?
🎓
Because the "normal" affinity is very different per target. GPCRs see endogenous ligands at pM-nM, so good drugs aim for similar potency. Enzyme inhibitors fight ATP or substrate so IC50 nM-μM is realistic. Ion channels (Nav1.7, Cav1.2) have state-dependent binding, where open and inactivated states differ by orders of magnitude. Nuclear receptors (ER, PR, GR) have deep steroid pockets, so pM-nM affinity is the norm. If your hit is 100x weaker than the class baseline, that is a big opportunity for medicinal-chemistry optimisation.
FAQ
The Hill equation Y = B + (T-B)/(1+(IC50/[D])^n) describes the relationship between dose and response, while IC50 and EC50 are parameters of that curve representing the concentration that gives half-maximal response. IC50 (half maximal inhibitory concentration) is the inhibitor concentration that achieves 50% inhibition, and EC50 (half maximal effective concentration) is the agonist concentration that produces 50% of the maximum effect. The Hill coefficient n is the slope, with n=1 indicating independent binding, n>1 positive cooperativity, and n<1 negative cooperativity.
pIC50 is defined as -log10(IC50 [M]). In drug discovery, IC50 values span more than three orders of magnitude (nM to μM), so raw numbers are hard to compare. pIC50=7 corresponds to IC50 of 100 nM and pIC50=9 to 1 nM. Each integer increase corresponds to a tenfold gain in potency, which is intuitive. SAR discussions, papers and patents almost always use pIC50.
When n drifts from 1, suspect the assay first. n much greater than 1 (e.g. 3+) often indicates receptor aggregation, multiple binding sites, non-specific adsorption, compound solubility limits or detector saturation. n much less than 1 (0.5 or below) suggests a heterogeneous receptor population, mixed mechanisms or compound degradation. True cooperativity (e.g. haemoglobin n≈2.8) needs structural or mechanistic backing. In day-to-day pharmacology n=1±0.3 is considered a healthy fit.
Receptor architecture, ligand size and binding-pocket depth set the typical affinity. GPCRs (β2-adrenergic, opioid, etc.) bind endogenous ligands at pM-nM, so drug EC50s aim for the same range. Enzymes (kinases, COX, ACE) compete with ATP or substrate, so IC50 of nM-μM is normal. Ion channels (Nav1.7, Cav1.2) show state-dependent binding and span a wide range, while nuclear receptors (ER, PR, GR) bind steroid-like ligands in deep pockets and frequently sit at pM-nM.
Real-world applications
Hit and lead optimisation in drug discovery: A typical workflow screens ~100k compounds via HTS, picks the IC50 < 1 μM hits, and uses medicinal-chemistry cycles to push potency into the nM range. Eleven-point dilution dose-responses are run on FlexStation, PathHunter, HTRF or Octet, and the data are fit with the four-parameter Hill equation in GraphPad Prism or Dotmatics Vortex.
Regulatory dossiers (FDA, PMDA, EMA): Dose-response curves are mandatory in pre-clinical packages. FDA Pharmacology/Toxicology Reviews, PMDA CTD Module 4, and EMA Module 4 evaluate in vitro IC50/EC50 and off-target selectivity. A selectivity ratio of off-target IC50 / on-target IC50 above ~100 is a common safety benchmark for clinical candidates.
Receptor-class baselines: GPCRs (β2-adrenergic, μ-opioid, dopamine D2) usually sit at EC50 0.1 nM-1 μM. Enzymes (EGFR kinase, COX-2, ACE) typically span IC50 1 nM-10 μM. Ion channels (Nav1.7, Cav1.2, hERG) cover IC50 1 nM-100 μM, and nuclear receptors (ER, PR, GR) reach EC50 0.01 nM-1 μM. The receptor selector in this tool surfaces those baselines.
Tox and hERG screening: A common heuristic is hERG IC50 / efficacy IC50 ≥ 30. Drop below that and the risk of torsades de pointes rises sharply. The same Hill equation is used to fit the hERG curve so that the selectivity window can be reported alongside the on-target potency.
Common pitfalls
The biggest trap is fitting with Top and Bottom fixed when the data do not actually span 0% and 100%. If your dilution range stops short of saturating the upper or lower plateau, a four-parameter fit returns wildly uncertain IC50 values. As a rule of thumb you want at least 8-10 points spanning two log units on either side of IC50; if you cannot, fix Top or Bottom (e.g. Bottom = 0) and report it explicitly as a constrained fit.
Second, blindly trusting an unconstrained Hill coefficient. If the fit returns n = 2.5 or n = 0.4, suspect the assay before celebrating cooperativity. Compound out of solubility, DMSO too high, detector saturated, mixed mechanisms - all common culprits. Always inspect the raw points first and rule out technical artefacts before publishing the n value.
Third, extrapolating an in vitro IC50 straight to a clinical dose. Cell-based or biochemical IC50 ignores plasma protein binding (PPB), tissue distribution, metabolism, active metabolites and blood-brain-barrier transit. To predict an effective plasma concentration you need free-drug levels, PK/PD modelling or PBPK. An IC50 of 10 nM with 99% PPB leaves a free concentration far below 10 nM, often pushing target plasma into μM territory.
How to Use
Enter IC50 value (nM) in the input field—this is the concentration producing 50% inhibition of your target enzyme or receptor.
Set Hill coefficient (typically 0.8–2.5 for most drugs); values >1 indicate positive cooperativity, <1 indicate negative cooperativity.
Adjust top response (maximum effect, usually 100%) and bottom response (baseline, usually 0%) to match your assay conditions.
Input a test dose concentration to calculate real-time response percentage using the Hill equation: Response = Bottom + (Top−Bottom)/(1+(Dose/IC50)^−n), where n is the Hill coefficient.
Read output metrics including pIC50 (−log IC50), IC10/IC90 concentration thresholds, and therapeutic ratio.
Worked Example
For a kinase inhibitor screening against Aurora-B kinase: IC50=45 nM, Hill coefficient=1.2, top response=100%, bottom response=2%. At a test dose of 150 nM, the simulator calculates Response=92.3%. The derived pIC50=7.35, IC10=3.8 nM, IC90=532 nM, yielding therapeutic ratio=140. A compound with Hill coefficient >1.0 exhibits sigmoidal binding kinetics, desirable for selectivity windows in multi-target profiling.
Practical Notes
Hill coefficient significantly affects the steepness of your dose-response curve; recombinant p38 MAPK often exhibits n=0.9–1.1, while allosteric modulators frequently show n>2.0, demanding careful interpretation of potency shifts.
IC50 shifts between assay platforms (biochemical vs. cellular) are common—kinase assays in ATP-competitive format may yield 5–15 fold differences versus cell-based IC50 due to target engagement and cell penetration variability.
The IC90/IC10 therapeutic ratio quantifies selectivity window; ratios below 10 indicate steep kinetics unsuitable for dose escalation in clinical settings, while ratios >50 enable flexible dosing strategies.
pIC50 conversion (pIC50=−log IC50 in M) standardizes potency reporting across literature; a 50 nM inhibitor=pIC50 7.30, enabling rapid literature benchmarking.