Detection of Velocity and Diffusion Coefficient Change Points in Single-Particle Trajectories
CELL PRESS. 2018: 217–29
The position-time trajectory of a biological subject moving in a complex environment contains rich information about how it interacts with the local setting. Whether the subject be an animal or an intracellular endosomal vesicle, the two primary modes of biological locomotion are directional movement and random walk, respectively characterized by velocity and diffusion coefficient. This contribution introduces a method to quantitatively divide a single-particle trajectory into segments that exhibit changes in the diffusion coefficient, velocity, or both. With the determination of these two physical parameters given by the maximum likelihood estimators, the relative precisions are given as explicit functions of the number of data points and total trajectory time. The method is based on rigorous statistical tests and does not require any presumed kinetics scheme. Results of extensive characterizations, extensions to 2D and 3D trajectories, and applications to common scenarios are also discussed.
View details for DOI 10.1016/j.bpj.2017.11.008
View details for Web of Science ID 000438958800009
View details for PubMedID 29241585
View details for PubMedCentralID PMC6051264
Parallelization of Change Point Detection
JOURNAL OF PHYSICAL CHEMISTRY A
2017; 121 (27): 5100–5109
The change point detection method ( Watkins , L. P. ; Yang , H. J. Phys. Chem. B 2005 , 109 , 617 ) allows the objective identification and isolation of abrupt changes along a data series. Because this method is grounded in statistical tests, it is particularly powerful for probing complex and noisy signals without artificially imposing a kinetics model. The original algorithm, however, has a time complexity of [Formula: see text], where N is the size of the data and is, therefore, limited in its scalability. This paper puts forth a parallelization of change point detection to address these time and memory constraints. This parallelization method was evaluated by applying it to changes in the mean of Gaussian-distributed data and found that time decreases superlinearly with respect to the number of processes (i.e., parallelization with two processes takes less than half of the time of one process). Moreover, there was minimal reduction in detection power. These results suggest that our parallelization algorithm is a viable scheme that can be implemented for other change point detection methods.
View details for DOI 10.1021/acs.jpca.7b04378
View details for Web of Science ID 000405761800004
View details for PubMedID 28616980