Accurate prediction of neutralization capacity from deep mining of SARS-CoV-2 serology

摘要 该项目的目标是建立一种准确和灵敏的预测中和的方法。 通过深入挖掘抗体谱来检测血清样本对抗SARS-CoV-2的能力。新冠肺炎大流行 仍然是一个全球威胁，有近700万病例和40万人死亡。在缺乏有效疫苗的情况下 而治疗方面，对SARS-CoV-2的免疫是预防SARS-CoV-2的主要机制 (再次)感染。我们最近对恢复期血清样本的研究表明，他们的中和水平 容量差异很大(超过100倍)，只有一小部分具有高中和能力。因为病毒 中和分析本质上是低吞吐量的，将其应用于高危人群是不现实的，例如 医院工作人员要及时上岗。不幸的是，两者之间只有适度的相关性 用标准的酶联免疫吸附试验测定中和能力和抗SARS-CoV-2抗体水平。 显然，我们仍然不知道哪些类型的抗体有助于病毒中和。我们最重要的是 在这个项目中要检验的假设是通过更深入地检查患者血清中的抗体图谱 在抗原、表位和抗体类型的定量方面，我们将能够识别出 病毒中和的预测性标志物。为此，我们将建立SARS-CoV-2的多重检测方法 血清学，这将使我们能够深入表征抗体图谱。然后我们将制定一个预测性的 算法的实现。我们已经组建了一支在抗体方面具有真正互补技能的专家团队 特征、病毒学和数据挖掘。我们有大量的康复血清样本， 这将使我们能够批判性地验证我们的技术。我们将尽快实现以下目标。(1) 我们将建立SARS-CoV-2的多重血清学检测方法，可分析多达15种抗体-抗原 在单一反应中的相互作用。主要的技术创新是引入了多维流动 细胞学。我们将生产包括Spike、受体结合区和核衣壳在内的多种抗原 蛋白质，以及它们的天然和设计的变体。我们将使用一个大的小组来改进和验证检测 恢复期血清样本。(2)我们将开发一种改进的病毒中和检测方法，以更好地量化 中和能力。(3)我们将开发一种预测中和能力的算法，该算法利用 我们多重化验的抗体图谱。这项分析将确定有助于 中和。该项目的最终产品将包括一种高通量的血清学分析，它将给出远- 比当前标准更丰富的抗体图谱，并伴随着准确的预测算法。一起， 这个平台将有助于加深对SARS-CoV-2感染的基本了解，以及 针对这种可怕病原体的疫苗和治疗方法的发展。

ABSTRACT The goal of this project is to establish an accurate and sensitive method for predicting the neutralization capacity against SARS-CoV-2 of serum samples by deep mining of antibody profiles. The COVID-19 pandemic remains a global threat with nearly seven million cases and 400K deaths. In the absence of effective vaccines and therapeutics, immunity against SARS-CoV-2 is a main mechanism of protection against SARS-CoV-2 (re)infection. Our recent studies of convalescent serum samples revealed that their levels of neutralization capacity vary greatly (over 100-fold) and only a small subset has high neutralization capacity. Because viral neutralization assays are inherently low throughput, it is unrealistic to apply it to a high-risk population such as hospital workers in a timely manner. Unfortunately, there is only moderate correlation between the neutralization capacity and the level of anti-SARS-CoV-2 antibody levels determined using standard ELISA. Clearly, we still do not understand what types of antibodies contribute to viral neutralization. Our overarching hypothesis to be tested in this project is that by examining the antibody profile in patient serum more deeply and quantitatively in terms of antigens, epitopes and antibody types, we will be able to identify quantitative predictive markers for viral neutralization. To this end, we will develop multiplex assay for SARS-CoV-2 serology that will enable us to deeply characterize the antibody profile. We will then develop a predictive algorithm by utilizing. We have assembled a team of experts with truly complementary skills in antibody characterization, virology and data mining. We have access to a large number of convalescent serum samples, which will enable us to critically validate our technology. We will expeditiously execute the following aims. (1) We will develop multiplex serology assay for SARS-CoV-2 that can profile up to 15 antibody-antigen interactions in a single reaction. The main technical innovation is the introduction of multi-dimensional flow cytometry. We will produce multiple antigens including Spike, receptor-binding domain and nucleocapsid protein, and their natural and designed variants. We will refine and validate the assay using a large panel of convalescent serum samples. (2) We will develop an improved viral neutralization assay to better quantify the neutralization capacity. (3) We will develop a predictive algorithm for neutralization capacity that utilizes the antibody profiles from our multiplex assay. This analysis will identify serology parameters that contribute to neutralization. The end products of this project will include a high-throughput serology assay that gives far- richer antibody profiles than the current standard accompanied with an accurate predictive algorithm. Together, this platform will help advance a fundamental understanding of SARS-CoV-2 infection as well as the development of vaccines and therapeutics against this formidable pathogen.