Stochastic, discrete modelling of in vitro and in vivo virus infections

背景：虽然实验本身推动了病毒学的进步，但它也有其局限性。即使在相对可控的体外细胞培养环境中，病毒感染也是一个复杂的、动态的系统。在实验上扰乱一种机制可能会引发其他几种机制，这使得仅凭观察就很难分离出其功能。病毒学通常也很紧迫(需要抗病毒药物和疫苗)，使解决方案驱动的研究成为优先事项，而不是建立对基本细胞-病毒相互作用的理解。我们在病毒物理领域的研究旨在通过对病毒感染期间发挥作用的关键机制进行定量描述(即数学和/或计算机模型，以下简称MCM)来填补这些空白。我们的工作就其描述和方法而言是物理学(应该这样评价)，但它所描述的系统本质上是病毒学的。方法：我们的MCM将细胞和病毒粒子(病毒粒子)的数量表示为连续变量，例如一个病毒粒子感染0.2个细胞，并描述实际上是离散的随机(DS)系统的确定性、平均场(MF)行为：一个病毒粒子感染一个细胞(或不感染)20%(或80%)的时间。拟议的工作建立在我们良好建立和验证的MF-MCM框架的基础上，总体目标是通过Rational、MCM辅助的方法来改进实验设计和结果解释。目的：开发和应用我们的MF-MCMS的DS版本，以确定和纠正日常病毒学中MF假设的错误：评估和改进旨在计数传染性病毒粒子的实验；解决样本中感染病毒粒子的数量与其可能导致的感染数量的关系；确定病毒粒子感染失败和部分流产感染如何影响共感染细胞中的病毒产量；以及探索呼吸道病毒感染过程中的随机事件，如咳嗽，如何改变其严重性和在人类呼吸道内传播的模式。更广泛地说，通过这些目标，我们的目标是改进病毒学实验数据的解释、统计分析和可视化。影响：这项工作可能会改变MCM使用和解释在病毒感染过程中通过实验测量的感染性病毒浓度的方式。区分具有高病毒粒子生成率和低病毒粒子细胞感染率的病毒株的能力可以告诉我们：什么共同的复制过程可能将共享一种策略的所有病毒连接在另一种策略之上；或者抗病毒药物应该针对的复制瓶颈。它可能从根本上改变我们看待、量化和比较病毒株适合性或抗病毒干预的有效性和作用模式的方式。

Context: While experimentation alone drives advances in virology, it has its limitations. Even within the relatively well-controlled environment of an in vitro cell culture, a virus infection is a complex, dynamical system. Disrupting a single mechanism experimentally can trigger several others, making it hard to isolate its function based on observations alone. There is also often an urgency to virology (need for antivirals, vaccines), making solution-driven research a priority over, and at the expense of, building an understanding of basic cell-virus interactions. Our research in the field of virophysics aims to fill these gaps through developing quantitative descriptions (i.e. mathematical and/or computer models, hereafter MCMs) of the key mechanisms at play during a virus infection. Our work is physics in its description and approach (and should be evaluated as such), but the systems it describes are virological in nature. Approach: Our MCMs represent the number of cells and virus particles (virions) as continuous variables, e.g. one virion infects 0.2 cells, and describe the deterministic, mean-field (MF) behaviour of what is in reality a discrete, stochastic (DS) system: one virion infects one cell (or not) 20% (or 80%) of the time. The proposed work builds on our well-established and validated MF-MCM frameworks with the overall goal to improve experimental design and result interpretation through rational, MCM-aided approaches. Objectives: To develop and apply a DS version of our MF-MCMs to identify and correct failures of the MF assumption in everyday virology to: evaluate and improve experiments meant to count infectious virions; resolve how the number of infectious virions in a sample relates to the number of infections it can cause; determine how virion infection failures and partial, abortive infections impact virus yield in co-infected cells; and explore how stochastic events over the course of a respiratory virus infection, such as coughing, alter its severity and pattern of spread within the human respiratory tract. More generally, through these objectives we aim to improve interpretation, statistical analysis, and visualization of experimental data in virology. Impact: This work could alter how MCMs use and interpret infectious virus concentrations measured experimentally over the time course of a virus infection. The ability to distinguish between a virus strain with a high rate of virion production and a low rate of cell infection per virion, and the opposite, can tell us about: what common replication process(es) might link all viruses that share one strategy over the other; or the replication bottleneck that should be targeted by antivirals. It could fundamentally alter how we view, quantify, and compare virus strain fitness or the efficacy and mode of action of antiviral interventions.