Computational Modeling Insights into Extreme Heterogeneity in COVID-19 Nasal Swab Data

Area of Science:

• Computational biology
• Virology
• Infectious disease modeling

Background:

• The COVID-19 pandemic generated extensive nasal swab data, revealing significant heterogeneity in viral loads across individuals.
• Understanding the drivers of this heterogeneity is crucial for comprehending viral respiratory infections.

Purpose of the Study:

• To investigate the factors contributing to the wide variation in viral loads observed in early nasal infections.
• To determine the impact of host physiology, viral kinetics, and testing timing on infection dynamics.

Main Methods:

• Development of a physiologically faithful, spatially resolved, stochastic computational model of nasal tract infection.
• Simulation of a "virtual population" with variations in host factors (mucus thickness, clearance) and viral kinetics.
• Application of global data analysis to isolate the sensitivity of infection outcomes to individual model parameters.

Main Results:

• Model simulations successfully reproduced the observed six-orders-of-magnitude heterogeneity in viral titers.
• The latency time of infected cells (time to viral RNA shedding) was identified as the dominant, exponential driver of population heterogeneity.
• Viral RNA shedding rate and the timing of nasal swab tests were also significant contributors to heterogeneity.

Conclusions:

• Early nasal infection dynamics and population heterogeneity are most strongly influenced by the time from initial infection to viral RNA shedding.
• Host-specific factors and viral characteristics play a role, but the timing of sample collection is a critical determinant of observed viral loads.
• The findings provide insights into the generalizability of SARS-CoV-2 infection dynamics to other viral respiratory infections.