Influenza is a high impact respiratory pathogen that affects human health worldwide. Influenza is associated with significant morbidity and mortality, with 30,000 to 40,000 annual deaths in the United States. Vaccination remains the primary method of influenza prevention. However, despite the overall public health success of annual influenza vaccinations, many individuals fail to induce a significant antibody response. Impaired vaccine responses are a particular issue in older adults, with estimates of efficacy around 50% and worse. Improved understanding of the biological mechanisms that influence the immune response to vaccination may offer clues on novel vaccine candidates and vaccination strategies. Systems vaccinology studies combine high-throughput experimental profiling techniques with computational analysis to provide an integrated, dynamic view of vaccine- driven immune responses. The validity of such studies is strengthened by integration of multiple studies, as single study cohort sizes are generally small and results vary due to differences in populations, experimental procedures, and other batch effects. Influenza vaccination studies comprise nearly 20% of the non-clinical trial studies currently available in the NIH/NIAID ImmPort repository and offer the potential for secondary analysis to identify robust signatures of vaccination responses that can guide improvements to vaccine efficacy. However, such integration is time-consuming, error-prone, and requires technical programming expertise that is not accessible to many researcher groups. The semantic web provides a theoretical and technical framework through which these problems can be addressed by leveraging ontology mappings to link related data and achieve FAIR data principles (Findability, Accessibility, Interoperability and Reusability). Here we propose to leverage the semantic web framework to link ImmPort vaccination studies to each other, and to the myriad of external public resources so critical for systems vaccinology (e.g., pathway databases, gene ontology annotations and publications). The resulting system (LinkedImm) will include public interfaces for hypothesis- based queries, and will be applied in a multi-study analysis to identify robust temporal signatures of the influenza vaccination response. We propose to achieve this goal through two aims:
(Aim 1) Leverage semantic web technologies to integrate influenza vaccination studies in ImmPort and link them with public resources to allow hypothesis-driven queries.
(Aim 2) Identify signatures of human influenza vaccination responses through a multi- study analysis. In summary, we propose a combination of technology-driven resource creation to facilitate ImmPort data reuse (Aim 1) with a driving biological project (Aim 2) of scientific and medical importance.
Despite the clear overall public health success of annual influenza vaccinations, many individuals fail to induce a significant antibody response following vaccination. This project will integrate the large number of vaccination studies available in the NIH/NIAID ImmPort data repository to carry out a multi-study analysis that will identify robust transcriptional signatures associated with successful (and unsuccessful) influenza vaccination responses. Further secondary analysis by the wider scientific community will be facilitated through a framework built on semantic web technologies that will enable automatic execution of hypothesis-based queries that span multiple ImmPort studies and integrate public resources through the linked open data cloud.
