Why You Should Never Manually Compensate Your Data
Written by Tim Bushnell, PhD / Figures courtesy of Pratip K. Chattopadhyay, PhD
One of the most important steps in proper flow cytometry is the process of compensation.
There are a lot of rumors and mysteries that fill laboratory notebooks about the process. Some of these processes are correct while others lead to incorrect compensation, resulting in poor data.
Compensation is the process for correcting for the spillover.
Spillover is the overlap of a fluorochrome into a second channel due to the physics of fluorescence. This is a mathematical value that ensures that contributions from the fluorochromes not being displayed are not affecting the distribution of the data being displayed.
One of the most common rumors or practices that has been passed down incorrectly by word of mouth over years past is the concept of manual compensation.
Manual compensation is the process of adjusting the compensation based on how the data visually looks.
If you have manually compensated data in your lab notebook–strike it out now.
Manual compensation results in overcompensated data, yielding incorrect conclusions. The best practice is to use automatic compensation algorithms that are available in the most current versions of flow cytometry software that are based on the work by Bagwell and Adam (1993) Ann NY Acad Sci 20:167-184, who describe the mathematics behind multiparameter compensation.
Keys To Automatic Compensation
For automatic compensation to be successful, the following three rules must be followed.
1. Controls must be at least as bright as the samples they will be applied to. Brighter is better, but not off scale.
2. Background fluorescence should be the same between the negative and positive population. Avoid using the universal negative for compensation.
3. The compensation color must be the same as the experimental color. For example, don’t use Alexa488 to compensate for FITC.
In every case where investigators have had “problems” with compensation, it is because they have violated the above three rules.
Now if these three rules have been followed, and the compensation doesn’t look right, do not make the mistake of editing the compensation matrix to make the data look better.
As many as 30% or researchers “tweak” the matrix to make it look better.
Follow the above three rules and your compensation will be correct and if you have issues, explore what those problems are and work to resolve them rather than making up fiction by manual compensation.
To learn more about automatic compensation and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on Mastery Class wait list.