Collaborative Research: A Molecular-to-Continuum, Data-Driven Strategy for Mucus Transport Modeling

该项目为肺部气道液体的流动和吸入颗粒（病原体、颗粒、药物载体颗粒）沉积在人体气道中的扩散开发预测数学理论和计算工具。在实验数据的基础上，开发了预测数学模型和计算工具。本文收集了肺培养和临床患者支气管上皮黏液中颗粒在黏液中的扩散和控制黏液流动的物理性质的实验数据。这种实验-理论-计算策略有望直接应用于各种肺部疾病和疾病的临床治疗，既可以用于疾病评估，也可以用于物理治疗和药物治疗策略的设计。在数学上，这些模型和模拟工具有望提供对生理强迫（呼吸和咳嗽产生的纤毛和空气阻力）的平均粘液流动特性的见解，并解决粘液分子网络中的微观结构变化。这些工具可以深入了解疾病和疾病进展期间粘液的生物物理差异，这是物理和药物治疗预测设计的关键。当与临床知识相结合时，该模型将提供推断黏液样本的流动和扩散运输特性以及健康和疾病状况的能力，并能够根据患者的个体化情况测试恢复黏液清除的疗法。本项目开发了一种数据驱动的肺粘液运输建模策略，将随机分子动力学过程、基于微观结构的应力的进化方程和流动的动量方程联系起来。实验数据包括粘液微观结构的随机（熵波动）和确定性（受控强迫）探针，以及来自人肺组织的细胞培养液中粘液运输的高分辨率显微镜。这些丰富的数据集提供了前所未有的由单个黏液蛋白分子、黏液凝胶中的缠结网络、瞬态黏液蛋白交联和链断裂动力学引起的广泛黏液弛豫谱的探针。细胞培养提供了对纤毛驱动的粘液流动运输的见解，并提供了一个实验室环境来探索施加物理压力和化学剂量的分子到宏观后果。为了将这些显著的数据转化为对粘液运输的预测性理解，研究了一个建模平台，该平台可以解决粘液的分子到连续体过程，无论是接近还是远离平衡。我们的策略从各种浓度黏液凝胶中的被动微珠探针的随机时间序列开始，以解决线性（接近平衡）粘弹性表征的逆和直接问题。接下来，在可控磁力范围内使用相同浓度的活性微珠数据来确定非线性阈值和微尺度下非平衡行为的特征。这些数据直接提供给这个项目的主要目标，这是制定一个新的微观-宏观本构律的粘液。提出这个公式是为了解释线性和非线性数据，并将分子到连续体的过程整合到一个新的粘液运输模型中。目前正在研究一种直接数值模拟的数值策略，并与每种实验的数据进行比较。

The project develops predictive mathematical theory and computational tools for both the flow of lung airway liquids and the diffusion of inhaled particles (pathogens, particulates, drug carrier particles) that are deposited in human airways. The predictive mathematical modeling and computational tools are developed on the basis of experimental data. The experimental data on particle diffusion in mucus and the physical properties that govern mucus flow transport are collected on human bronchial epithelial mucus from lung cultures and clinical patients. This experimental-theoretical-computational strategy has the promise for direct applications in clinical treatment of humans with diverse lung diseases and disorders, both for assessment of disease and for design of physical therapy and drug treatment strategies. Mathematically, these models and simulation tools promise to provide insight into mean mucus flow properties from physiological forcing (cilia and air drag from breathing and cough) and also to resolve microstructural changes in the mucus molecular network. These tools can offer insight into the biophysical differences in mucus during disease and disease progression, which are keys to a predictive design of physical and drug therapies. When integrated with clinical knowledge, this modeling will provide the capability to infer flow and diffusive transport properties of mucus samples and healthy and disease conditions, and the capability to test therapies to reinstate mucus clearance on an individualized patient basis.This project develops a data-driven strategy for the modeling of lung mucus transport, linking stochastic molecular kinetic processes, evolution equations for microstructure-based stresses, and momentum equations for flow. The experimental data consists of stochastic (entropic fluctuations) and deterministic (controlled forcing) probes of mucus microstructure, together with high-resolution microscopy of mucus transport in cell cultures derived from human lung tissue. These rich data sets provide unprecedented probes of the broad mucus relaxation spectrum arising from single mucin molecules, their entanglement network in mucus gels, transient mucin crosslinking, and chain scission kinetics. Cell cultures afford insights into cilia-driven flow transport of mucus, and provide a laboratory setting to explore molecular-to-macroscopic consequences of imposed physical stresses and chemical dosing. To translate this remarkable data into a predictive understanding of mucus transport, a modeling platform is studied that resolves molecular-to-continuum processes of mucus, both near and far from equilibrium. Our strategy begins with stochastic time series of passive microbead probes in mucus gels of various concentrations, to solve the inverse and direct problems for linear (near equilibrium) viscoelastic characterization. Next, active microbead data for the same set of concentrations is used over a range of controlled magnetic forces to determine nonlinear thresholds and the signatures of non-equilibrium behavior at the microscale. These data directly feed into the main objective of this project, which is the formulation of a new microscopic-macroscopic constitutive law for mucus. This formulation is proposed to interpret the linear and nonlinear data, and to integrate molecular-to-continuum processes into a new mucus transport model. A numerical strategy is under study for direct numerical simulations and comparison with data from each type of experiment.