Computational Modeling and Simulation of Viral Load Kinetics in SARS-CoV-2 Replication

Abstract

COVID-19, which has been attracting more and more attention worldwide, reactivates in some recovery cases. In this study, two research questions related to reactivation are proposed and addressed to compare the reactivation risk between mild and severe cases. The first one is about the time lag between when the symptoms disappear and when the viral load declines. The second one is about how long post-cure asymptomatic duration continues and the extent to which course is involved in the severity of the disease. We construct an expanded computational model for describing viral load kinetics in SARS-CoV-2 replication and investigate the solution trajectory’s behavior in the phase space to clarify the relationship among variables. Following this, we implement a global sensitivity analysis of the model to specify its robustness. Based on the sensitivity analysis results, we fit the model to empirical data of the viral load in mild cases and severe cases cited from one previous research. We check the data property by data visualization. As a result, we observe the differences in the duration of the period without symptoms and the convergence timing corresponding to the disease’s severity, reflecting that the asymptomatic carriers in severe cases would be at higher reactivation risk exposed to longer infectious duration. Moreover, we reproduce the lower mortality of infected cells by viral immune evasion mechanisms in severe cases. Although this study cannot suggest a quantitative implication due to a lack of extensive observed data, it successfully shows the viral load kinetics and the qualitative relationship between the severity of COVID-19 and the reactivation risk of asymptomatic carriers.