Saturation:Analysis of Data:Nonlinear Regression Analysis
Explanation of Nonlinear Regression Analysis
The saturation curve has the shape of a hyperbola. The following equation applies to this curve
Where B is Bound and F is Free
Nonlinear regression analysis uses this equation to obtain estimates of Kd
and Bmax. It then determines how well the data points fit the equation as
illustrated here. 
Return to "nonlinear regression analysis of sample
experiment" under Plotting of Sample Experiment
| In nonlinear regression analysis initial estimates of the Kd and Bmax values are made. From the graph of the sample experiment the estimate of Bmax was
17 pM |
 |
 |
| These values are entered into the equation for a hyperbola to generate a line. |  |
|
| The distance between the line and each of the points is determined. These distances are squared and summed (sum of the squares). |  |
|
| A new estimate of Kd and Bmax are made. Since the initial values were too low, higher values will be estimated this time Bmax = 18 , Kd= 250 These values are entered into the equation for a hyperbola to generate a line. |
 |
|
| The sum of the squares of the distance between this new line and the data points are determined. The sum of the squares from this iteration is compared to the sum of the squares from the previous iteration. |
 |
|
| Further estimates of Kd and Bmax are made and entered into the equation for the hyperbola. The sum of the squares are determined. This process continues until there is no further change in the sum of the squares. The last estimates of Kd and Bmax are considered to provide the best fit. This can be a rather time consuming process and is usually accomplished using a computer program such as Prism sold by GraphPad Inc. |
|
 |