How Should Eukaryotic Chemotaxis be Measured?

Chemotaxis and directed cell migration are fundamentally important to eukaryotic biology. To understand the mechanisms that drive such complex processes, informative and robust measurements are essential, but the field does not always agree what these should be. Here we identify the most dependable measures of chemotactic steering and the underlying cell migration, and provide tools to test them. Some widely-used metrics can end up misleading, in particular "cos 8" for directional accuracy. We suggest that chemotactic efficiency should be used as a primary metric. Mean squared displacement and directional autocorrelation can be used to unpick different models of random and directed cell migration. Transition matrices are another useful tool for understanding migration mechanisms and avoiding artefacts, and provide a graphical illustration of how well cells maintain each direction. Unexpectedly, the choice of start and end points of tracks strongly affects the measurements and can seriously bias the measured results. This is particularly clear when cells are not homogeneously distributed at the start of an assay. To support straightforward adoption of these methods, we provide a suite of tools as a plugin for the open-source ImageJ program, and describe how they can be used to understand complex scenarios like self-generated chemotactic gradients.