If you don’t log-transform, Prism will fit a hyperbola (not sigmoid) and fail.

| Parameter | Report as | |-----------|------------| | EC50 | 1.2e-7 M (95% CI: 8.5e-8 to 1.7e-7) | | LogEC50 | -6.92 ± 0.08 | | Hill slope | 1.2 ± 0.15 | | Top | 98.3% ± 2.1 | | Bottom | 4.2% ± 1.3 |

(advanced):

| Problem | Solution | |---------|----------| | Curve looks like a straight line | X values too narrow range – use 5–10 log-spaced concentrations | | EC50 confidence intervals huge | Poor data at inflection point – add more points near EC50 | | Prism says “unstable parameters” | Too many parameters for data – use 3-parameter model | | Negative or >100% responses | Check baseline correction or outlier removal | | IC50 vs. EC50 confusion | IC50 = inhibition curve (descending); EC50 = activation (ascending) |

To create a molarity graph in GraphPad, follow these steps: