Normalization of data for viability and relative cell function curves

Abstract

Many types of assays in cell biology, pharmacology and toxicology generate data in which a parameter is measured in a reference system (negative control) and then also under conditions of increasing stress or drug exposure. To make such data easily comparable, they are normalized, i.e., the initial value of the system (e.g., viability or transport function) is set to 100%, and all data are indicated relative to this value. Then, curves are fitted through the data points and summary data of the system behavior are determined. For this, a benchmark response (BMR) is given (e.g., a curve drop by 15 or 50%), and the corresponding benchmark concentration (BMC15 or BMC50) is determined. Especially for low BMRs, this procedure is not very robust and often results in incorrect summary data. It is often neglected that a second normalization (re-normalization) is necessary to make the data suitable for curve fitting. It is also frequently overlooked that this requires knowledge of the system behavior at very low stress conditions. Here, good in vitro practice guidance for the re-normalization procedure is provided so that data of higher fidelity can be generated and presented.