Nonlinear regression with Prism

spa

s	Nonlin with Prism
	Initial values
	Fixing constants
	Method options
	Output options
	Default options
	Importing equations
	Writing equations
	Constraining
	Two models in one
	Simulate a curve

Interpreting the results

Comparing two curves

Distributions of best-fit values

curvefit.com was created by GraphPad Software, Inc. Send comments or questions to the author of these pages, Dr. Harvey Motulsky, president of GraphPad Software.

In April 2003, GraphPad released Prism 4 and published Fitting Models to Biological Data using Linear and Nonlinear Regression. This book includes all the information that comprises curvefit.com, and much more. You can read this book as a pdf file.

Nonlinear regression with Prism

Prism makes nonlinear regression very easy. Follow these steps. You may start either from a table of data or from a graph.

Step 1. Press the Analyze button and select built-in analyses.

Step 2. Select Nonlinear regression from the list of curves and regressions.

Step 3. Pick an equation. See Classic equations commonly used by biologists, Importing equations and the equation library, and User-defined equations. For general information on creating models, see How models are derived.

To compare two equations, select "Fit to two equations and compare with an F test." Check the option button for equation 1, and pick that equation. Then check the option button for equation 2 and pick the second equation. See Comparing the fits of two models.

Step 4. Select any additional calculations you wish to perform.

Calculation	Description
Calculate unknowns from a standard curve	After fitting a curve, Prism will interpolate unknown values from that curve. See Reading unknowns from standard curves .
Runs test	Tests whether the curve deviates systematically from your data. See Runs test
Residuals	Helps you test the assumptions of nonlinear regression. See Residuals.
Dose-ratios for Schild plot	Computes the log of dose-ratios from a series of dose-response curves. See Dose-response curves in the presence of antagonists.
Ki from IC50	After fitting competitive binding curves, calculate the Ki from the IC50. See Competitive binding data with one class of receptors
t test to compare best-fit variables	If you fit two data sets, compare the best-fit values of each parameter using a t test. See Comparing fits to two sets of data (same equation). The results will be in the "Overview and comparison" view (page).

Step 5. Set other options by clicking one of the four buttons at the bottom left of the nonlinear regression parameters dialog. Clicking these buttons will bring up a second dialog.

Option	Description
Initial values	Prism automatically supplies initial values for all variables, so you rarely have to think about them. Press this button to inspect and alter the initial values. See Initial values with Prism.
Constants	Fix the value of one or more parameters in the equation to a constant value. See Fixing parameters to constant values .
Method	Choose how Prism weights data, handles replicates, and other options. See Method options with Prism
Output	Configure the output by choosing the beginning and end of the curve, the number of significant digits, and more. See Output options with Prism.

Step 6. Press OK to fit the curve. Prism will superimpose the curve on a graph of your data, and will place the numerical results into the Results folder. For guidance interpreting the results, see Approach to interpreting nonlinear regression results.

Initial values with Prism