Modeling Statistical Multiplicity of Infection to Update Methods of Virus Quantification and Infectivity Assays

Many biological assays are employed in virological studies in order to quantify physical parameters of interest such as the number of viruses present in a solution or the ability of a viral strain to successfully infect a host cell. At the dilute concentrations that virus quantification assays operate, the results can be subject to the stochastic variability in the virus-cell interactions. At the other extreme, large numbers of virus particles are used in infectivity assays, resulting in a statistical multiplicity of infection (SMOI) where each cell is infected by multiple particles. In many cases, the SMOI may lead to significant variability in estimations of desired physical parameters. In our study, we develop probabilistic models for SMOI at low and high viral particle concentrations and apply them to several assays used in virological studies including plaque assays, endpoint dilution, and luciferase reporter assays. After performing a detailed analysis of statistical effects, we propose improved estimates for inferring experimental and biophysical parameters from the results of these assays.