Tracking particles is an incredibly challenging problem in scientific research. To illustrate the difficulty, tracking just 7 particles over 10 frames presents more unique configurations for how to connect particles between frames than there are atoms in the observable universe. It is virtually impossible to check every possible tracking configuration. Most tracking algorithms are tailored for a high number of particles and, as a result, use shortcuts to check only a small subset of configurations. This approach can lead to suboptimal tracking solutions, which are unavoidable when working with a large number of particles that need to be tracked simultaneously. However, when tracking fewer or more well-separated particles or when additional constraints can be imposed, we can afford to be more thorough in the search for the optimal tracking. In this direction, we developed ^★Track, an algorithm designed to track multiple replicating and persistent objects like DNA loci, while accounting for missing, merged, and spurious detections. ^★Track works by first converting the tracking problem into a path-finding problem on a layered graph. It then uses the A*-graph traversal algorithm to find the best set of trajectory segments. We then use the fact that the tracked objects are persistent objects (i.e., they do not disappear) to impose significant constraints on how these segments can be combined into a complete tracking. This innovative approach has led to interesting results on both chromosome and plasmid maintenance systems.

The cover image of the May 2 issue of Biophysical Journal shows the layered graph in the background, with a time series of five cross-sectioned bacterial cells (left to right) in the foreground. Inside each cell is one chromosome, upon which multiple plasmids are situated. The lines connecting the plasmids represent the optimal tracking found in the layered graph. This rendering was created by using the 3D modeling software Blender.

In summary, our study demonstrates the potential impact of highly accurate tracking when analyzing DNA loci. The ^★Track algorithm offers a powerful and versatile tool for uncovering new insights into various cellular processes and the maintenance of chromosome and plasmid systems. Remarkably, ^★Track is not limited to DNA loci and can be applied to a wide range of persistent objects in both prokaryotic and eukaryotic organisms, making it an invaluable resource for researchers across various fields. If you are interested in this algorithm or other software solutions that we developed, visit our group website at https://www.mpi-marburg.mpg.de/murray.

— Robin Köhler, Ismath Sadhir, and Seán M. Murray