Evaluation of bacterial motility from non-Gaussianity of finite-sample trajectories using the large deviation principle
Evaluation of bacterial motility from non-Gaussianity of finite-sample trajectories using the large deviation principle
复制标题
DOI：
10.1088/0953-8984/25/46/465103
复制
发表时间：
2013-11-20
影响因子：
2.7
通讯作者：
Kawano, Satoyuki
中科院分区：
文献类型：
--
作者：
Hanasaki, Itsuo;Kawano, Satoyuki
通常在轨迹数据中识别细菌的运动性，并将其与布朗运动进行比较，但扩散系数不足以评价它。本文提出了一种基于大偏差原理的方法。我们表明，它可以用来评估模型大肠杆菌运动的非高斯特性，并区分导致相同扩散系数的平均运行持续时间和运行速度的组合。我们提出的方法不需要化学刺激来诱导特定方向的趋化性，它适用于各种类型的自行车式运动，对于这些运动来说，例如奔跑和翻滚的阈值参数或头/尾方向的先验信息是不存在的。我们还讨论了大偏差量的有限样本效应问题，但我们建议利用它来刻画运动性的性质。
Motility of bacteria is usually recognized in the trajectory data and compared with Brownian motion, but the diffusion coefficient is insufficient to evaluate it. In this paper, we propose a method based on the large deviation principle. We show that it can be used to evaluate the non-Gaussian characteristics of model Escherichia coli motions and to distinguish combinations of the mean running duration and running speed that lead to the same diffusion coefficient. Our proposed method does not require chemical stimuli to induce the chemotaxis in a specific direction, and it is applicable to various types of self-propelling motions for which no a priori information of, for example, threshold parameters for run and tumble or head/tail direction is available. We also address the issue of the finite-sample effect on the large deviation quantities, but we propose to make use of it to characterize the nature of motility.