Improved estimation of anomalous diffusion exponents in single-particle tracking experiments
Improved estimation of anomalous diffusion exponents in single-particle tracking experiments
复制标题
DOI：
10.1103/physreve.87.052713
复制
发表时间：
2013-05-20
影响因子：
2.4
通讯作者：
Garini, Yuval
中科院分区：
文献类型：
--
作者：
Kepten, Eldad;Bronshtein, Irena;Garini, Yuval
The mean square displacement is a central tool in the analysis of single-particle tracking experiments, shedding light on various biophysical phenomena. Frequently, parameters are extracted by performing time averages on single-particle trajectories followed by ensemble averaging. This procedure, however, suffers from two systematic errors when applied to particles that perform anomalous diffusion. The first is significant at short-time lags and is induced by measurement errors. The second arises from the natural heterogeneity in biophysical systems. We show how to estimate and correct these two errors and improve the estimation of the anomalous parameters for the whole particle distribution. As a consequence, we manage to characterize ensembles of heterogeneous particles even for rather short and noisy measurements where regular time-averaged mean square displacement analysis fails. We apply this method to both simulations and in vivo measurements of telomere diffusion in 3T3 mouse embryonic fibroblast cells. The motion of telomeres is found to be subdiffusive with an average exponent constant in time. Individual telomere exponents are normally distributed around the average exponent. The proposed methodology has the potential to improve experimental accuracy while maintaining lower experimental costs and complexity.
