Cytomegalovirus (CMV) infects around 50% of the US population. Even though the CMV exists in a latent state in healthy individuals, it profoundly shapes the immune system. Recent studies suggest that the CMV infection alters the immune response to influenza vaccine. However, the exact effect of CMV on the efficacy of the influenza vaccine remains controversial. In addition, how CMV shapes the immune responses toward other vaccines are unknown. We hypothesize that latent CMV infection induces critical changes in the immune system, which alters the efficacy of multiple types of vaccines. The ImmPort database currently hosts 133 vaccine studies, covering 21 types of vaccines, creating an unprecedented opportunity for us to test our hypothesis. We will perform a comprehensive meta-analysis to test the relationship between CMV and vaccine efficacies, and will use state-of-art statistical models (e.g., Dynamic Bayesian Network) to identify the mechanism by which CMV alters the vaccine response. Leveraging the group's expertise in computational immunology and rich datasets on ImmPort, we will address the following aims.
Aim1 : Test the effect of CMV on influenza vaccine outcome. We will perform a meta-analysis of 60 influenza studies available on ImmPort to test the impact of CMV. We will quantify and standardize the efficacy of influenza vaccine across studies, which are measured by hemagglutinin inhibition (HAI) assays before and after the vaccination. We will also determine the CMV infection status in subjects, either directly from serological tests or indirectly from immune- phenotyping data using cutting-edge machine learning tools. We will then test if CMV increases the response to influenza vaccine by analyzing data from all studies in a unified statistical framework while taking the heterogeneity between studies into account.
Aim2 : Bayesian network analysis of influenza vaccine response. We will harmonize multimodal immune-phenotyping data from the influenza vaccine studies, including transcriptomics data, cytometry data, and cytokine measurements. We will use state-of-art network analysis methods (e.g., Dynamic Bayesian network) to model the interplay between the immune components over time. Using the Bayesian network, we will investigate the mechanism by which CMV shapes the outcome of influenza vaccination.
Aim3 : Explore the effect of CMV infection on other vaccines. We will extend our analysis to vaccines other than influenza vaccine, (e.g., West Nile, Hepatitis B, yellow fever, malaria, and Tuberculosis). We will quantify the vaccine efficacy using assays specific to the vaccine type, such as Controlled Human Malaria Infection (CHMI) for the malaria vaccine and Plaque Reduction Neutralization Test for the yellow fever vaccine. We will perform separate network analyses to characterize the relationship between CMV and the immune response of individual vaccines. We will then perform joint analysis across vaccine types to identify the common impact of CMV across vaccine types.
Infectious diseases remain an urgent problem, resulting in an estimated 3 million deaths worldwide and more than 110,000 deaths in the United States annually, but the efficacy of vaccines against many infectious diseases remains suboptimal, including malaria, influenza, and dengue. To improve the vaccines, it is crucial to understand the factors that affect the immune response toward vaccines. In this study, we investigate how cytomegalovirus affects the efficacy of multiple vaccines, providing valuable information for improving the design of vaccines.
