To calculate IC50, you would need a series of dose-response data (e.g., drug
concentrations x_{1}, x_{2},
...,x_{n} and growth inhibition y_{1}, y_{2},
...,y_{n}). The values of y are in the range of 0-1.
**Linear Regression**
The simplest estimate of IC50 is to plot x-y and fit the data with
a straight line (linear regression).
IC50 value is then estimated using the fitted line, i.e.,
Y = a * X + b,
IC50 = (0.5 - b)/a.
**Log transformation**
Frequently, linear regression is not a good fit to dose-response data.
The response-curve fits better to a straight line if the x-axis is
logarithm-transformed.
**Excel add-in**
ED50V10 (Readme) is an Excel add-in for
calculating IC50/EC50 values. Input your data in the left columns,
and your results will be shown in the right half of the Excel table.
To calculate IC50, input 50 in the "INTERPOLATE..." table (highlighted in blue),
the result will be shown on the right (highlighted in green).
For log-transformation, go to Data Transformation on the upper-right, input
3 in the DOSE (X-axis)/1st cell.
**Four-Parameter Logistic Function**
Four-parameter (A, B, C, D) logistic function (or Sigmoidal)
is frequently used to fit dose-response curves.
The drug data x is in logarithmic form. This is a standard function
in most statistics software.
The parameter C is the estimate of IC50 or EC50.
Table of Contents << 278 >> |