Determining the role of selection on the evolution of HIV-1 at the within- and between-host scales

HIV-1是一种RNA病毒，具有短世代时间，高突变率和慢性感染期，为快速适应性进化提供了理想的条件。进化动力学是由两个层面的选择塑造的：宿主内适应，如选择逃避宿主特异性免疫反应的变体，以及宿主间适应。过去的研究发现，宿主内的进化速度比宿主之间快五倍，但有证据表明，分歧的速度大致相当。对这一发现的一种解释是，隔离位点在感染期间在状态之间“切换”。可能的机制包括选择压力的时间变化和同义的轻微有害突变，由于遗传连锁，该项目将使用HIV-1测序数据的系统发育分析和群体遗传学方法来确定在状态之间切换的位点，并建立一个数据库，推动对在两个层面上塑造进化的选择过程的理解。选择压力对系统发育推断的影响目前在很大程度上是未知的，这项工作的一个关键目标将是开发包含进化复杂性的动态模型，如选择和重组，这些复杂性通常是未考虑的或未参考真实的生物学发现。发展一个突变-选择平衡模型，通过纯化选择解释毒力的进化.确定导致宿主内切换的选择过程：将分析纵向测序数据以确定选择中的位点和导致分离位点切换的最可能机制。将宿主内和宿主间水平的切换纳入进化和进化动力学模型：将解决的问题包括个体和群体水平的比率是否对测量时间尺度敏感，以及采样频率如何改变切换对系统发育推断的影响，例如到最近共同祖先的时间。新奇：这个博士的主要贡献将是生物机制的联系，选择压力以及宿主内和宿主间的进化和流行病学结果。有效的疫苗和治疗方法的开发取决于我们对病毒如何在具有不同免疫学背景的宿主之间传播以及因此多年来在宿主内进化和适应的理解，这在很大程度上受到不同选择过程的影响，这些选择过程是本研究的重点。此外，更复杂的生物学动态模型将改善对新爆发时间的推断，从而改善公共卫生干预措施的实施。EPSRC战略和研究领域的一致性该项目福尔斯属于EPSRC医疗保健技术研究领域，并将部分利用已经发表的深度测序数据集，以及将在6-12个月内提供的数百个个体和传播对的长读长纵向测序数据。

SummaryHIV-1 is an RNA virus with short generation times, high mutation rates and a chronic period of infection, providing the ideal conditions for rapid adaptive evolution. Evolutionary dynamics are shaped by selection on two levels: within-host adaptation, such as selection of variants that evade host-specific immune responses, and between-host adaptation. Past research has found evolutionary rates are five times faster within a host than between hosts, however there is evidence that suggests the rates of divergence are approximately equivalent. A proposed explanation of this finding is that segregating sites are "toggling" between states during infection. Possible mechanisms include temporal changes in selection pressures and synonymous slightly deleterious mutations hitch-hiking to high frequency due to genetic linkage, and then falling in frequency as a result of recombination or reversion.This project will use phylogenetic analysis of HIV-1 sequencing data and methods from population genetics to determine the sites toggling between states and to build a data-driven understanding of the selection process that shape evolution at both levels. The effect of selection pressures on phylogenetic inferences is currently largely unknown, and a key objective of this work will be to develop phylodynamic models that incorporate the evolutionary complexities, such as selection and recombination, that are typically unaccounted for or are included without reference to real biological findings.Aims and objectives:1. Develop a mutation-selection balance model that explains the evolution of virulence by purifying selection.2. Determine the selection processes that results in within-host toggling: longitudinal sequencing data will be analysed to identify sites under selection and the most likely mechanisms that result in toggling at segregating sites.3. Incorporate toggling at within and between-host levels into evolutionary and phylodynamic models: questions that will be addressed include whether individual and population-level rates are sensitive to the measurement timescale, and how does sampling frequency alter the effect of toggling on phylogenetic inferences, such as time to most recent common ancestor.Novelty:The major contribution of this DPhil will be the linking of biological mechanisms, selection pressures and evolutionary and epidemiological outcomes at the within and between-host levels. The development of effective vaccines and treatments depend upon our understanding of how the virus is transmitted between hosts with different immunological backgrounds and consequently evolves and adapts within the host over many years, and this is heavily influenced by the different selection processes that are the focus of this research. Moreover, more sophisticated phylodynamic models will improve the inferences made on the timing of new outbreaks and consequently the implementation of public health interventions.Alignment of EPSRC strategies and research areasThis project falls within the EPSRC healthcare technologies research area and will in part utilise already published deep-sequenced datasets, as well as long-read longitudinal sequencing data for hundreds of individuals and transmission pairs that will be available within 6-12 months.