After the COVID-19 pandemic, over 40 million children worldwide are at risk of measles due to delayed vaccination and temporary SARS-CoV-2 viral dominance. The lasting immunosuppression caused by the disease presents a major health threat, and treatment options are urgently needed, especially for low- and middle-income countries. The manuscript by Cox et al. (2024) explores features of canine distemper virus (CDV) in ferrets, using this model as a surrogate for measles to evaluate two possible antiviral treatments, ERDRP-0519 and GHP-88309. Ferrets were infected with a lethal challenge of CDV and treated with either drug or therapeutic vaccination. We aim to characterize both the infection dynamics and efficacy of the two drug treatments using the data from the PBMC (peripheral blood mononuclear cell) associated viremia titers of CDV infected ferrets and the lymphocyte counts measured during the duration of the study. A differential mathematical model was fitted to the experimental data by minimizing the sum of squared residuals (SSR), and errors in the parameter fits were estimated using Monte Carlo Markov Chain (MCMC). We visualized the key parameter distributions for each dataset using histograms, allowing us to directly compare how each treatment influences infection dynamics. The results revealed that ERDRP-0519 reduced viral entry and enhanced clearance while GHP-88309 improved target cell growth and increased the rate of infected cell death. These findings suggest that both drugs are potentially effective measles treatment options, with ERDRP-0519 having a direct antiviral effect and GHP-88309 aiding in immune recovery. Overall, these insights provide a foundation for optimizing treatment strategies and highlight the potential for both drugs to combat measles and related morbillivirus infections, especially in areas with limited resources and vaccines.

Gliomas account for approximately 27% of all primary central nervous system tumors and exhibit highly aggressive growth patterns, making conventional treatments ineffective. Previous research has demonstrated that a replication-competent Sindbis virus (SINV) combined with cytokines (IL-7, IL-12, and GM-CSF) shows promising results in slowing down glioma progression. While prior research demonstrated that SINV combined with cytokines reduces tumor growth, a quantitative understanding of its effects remains limited. This study aims to develop and fit a mathematical model of oncolytic virus infection to data from previous research to quantify key biological processes in glioma treatment. By parameterizing the Sindbis virus-glioma interaction and estimating the effects of cytokine therapy, this model aims to evaluate the efficacy of different SINV variants, with and without cytokine combinations, in controlling tumor growth. We use an Ordinary Differential Equation (ODE) model to describe tumor growth inhibition by the oncolytic SINVs. The model includes variables for uninfected and infected tumor cells, viral load, and cytokine concentration. The data extracted from published tumor growth curves will be used to estimate key parameters, including viral replication rate, tumor growth rate, and cytokine effects. Parameter fitting will be conducted by minimizing the Sum of Squared Residuals (SSR) between model predictions and experimental data. Error in the parameters will be estimated through bootstrapping to find the best fit parameters with 95% confidence intervals. Preliminary analysis suggests that the model effectively captures tumor growth rates observed in the experimental data. Parameter estimation provides insights into the viral infection rate, cytokine-induced tumor suppression, and the timing of viral injections. These findings will help refine our understanding of how the SINVs and cytokine therapy interact in glioma treatment. This study provides a quantitative framework for evaluating the therapeutic effects of an oncolytic SINV combined with cytokines in glioma treatment. By providing parameter estimates for key biological processes, our model can help optimize treatment strategies and guide future experimental research in oncolytic virotherapy.

Mathematical modeling of viral kinetics can be used to gain further insight into the viral replication cycle and virus-host interactions. However, many virus dynamics models do not incorporate the cell-to-cell heterogeneity of virus yield or the time-dependent factor of virus production. A recent study of the kinetics of the vesicular stomatitis virus (VSV) in single BHK cells determined that both the virus production rate and the yield of virus particles vary widely between individual cells of the same cell population. We used the results of this study to determine the distribution that best describes the time course of viral production within single cells. The best distribution was then used to incorporate time-varying production into a standard model of viral kinetics. The best-fit model was determined by fitting potential distributions to cumulative viral production from single cells and comparing the Akaike Information Criterion (AIC). The results show that the best fit for most cells was log-normal. Time-dependent viral production was modeled with an integro-differential equation that incorporated the log-normal probability distribution into a standard constant production model of viral kinetics. This time-dependent model was compared to one of constant production by examining the differences between the viral peak, time of the peak, upslope, downslope, and area under the curve. These findings could have further-reaching implications for helping define the time course and nature of a particular virus infection within the human body as well as improving the dose-timing and efficacy of anti-viral treatments.

Syncytia are multinucleated cells that can occur due to virus infection of cells. Mathematical models in the form of ordinary differential equations can be used to simulate the growth of syncytia. Several novel ODE models can explain syncytia growth. Before employing these models on actual data, it is essential to analyze their structural (theoretical) and practical identifiability using computer software. Structural identifiability is an inherent property of each model and its parameters, referring to our ability to determine parameter values for the model given particular experimental measurements. Practical Identifiability analysis of a model is concerned with determining our ability to accurately determine parameter values given experimental error. Combining these two techniques enables us to determine whether or not the parameters of our syncytia models can be accurately determined. Obtaining accurate parameter values allows us to make conclusions about our data that can provide insight into the nature of the spread of syncytia. From this, we can plan experiments to parameterize the syncytia growth in the contexts of our models.

Mpox virus is a type of virus similar to smallpox that can cause diseases in humans. Several experiments have been done to collect data on how mpox evolves within an infected host. This data can be analyzed within the context of mathematical models to determine important characteristics of mpox. From this analysis, we can estimate the growth rate, reproduction number, and infecting time of mpox.  We can also construct confidence intervals to estimate the error in our predictions using bootstrapping.  Bootstrapping allows us to analyze parameter correlations within mpox data to understand how parameter values within the model affect each other in our model. From these values and confidence intervals, we can learn about how mpox evolves within the body over time. This information, in turn, may allow us to make predictions on how mpox evolves within people during infection that could inform future treatment regimens.