Vaccination is arguably the most successful medical intervention against infectious diseases, yet some infectious viruses remain elusive. Rapid advancement in genetic sequencing technologies and immunological experiments are enabling rational designs of vaccines. Such next-generation vaccines can train the immune system to target specific fragments of the virus while avoiding others.

This thesis leverages data science approaches to guide the design of vaccines by identifying those specific fragments that the immune system of a diverse population can effectively target for three infectious viruses: human immunodeficiency virus (HIV), dengue virus, and severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2, which is the cause of COVID-19). Due to the underlying biology, these viruse...[ Read more ]

Vaccination is arguably the most successful medical intervention against infectious diseases, yet some infectious viruses remain elusive. Rapid advancement in genetic sequencing technologies and immunological experiments are enabling rational designs of vaccines. Such next-generation vaccines can train the immune system to target specific fragments of the virus while avoiding others.

This thesis leverages data science approaches to guide the design of vaccines by identifying those specific fragments that the immune system of a diverse population can effectively target for three infectious viruses: human immunodeficiency virus (HIV), dengue virus, and severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2, which is the cause of COVID-19). Due to the underlying biology, these viruses pose unique challenges to identifying the effective immune targets, therefore, by aggregating various types of data and developing and adapting different computational methods, we attempted to mitigate these challenges.

First, we improved the identification of an evolutionarily constrained region within an immunogenic protein of HIV, by principally modifying a spectral decomposition-based sectoring method that has been used to reveal mutational patterns in HIV. This work informs effective immunogen design for use in HIV vaccines. Second, we identified a set of vaccine targets that are potentially robust against multiple circulating subtypes of the dengue virus by performing a comprehensive conservation analysis. Third, for SARS-CoV-2, we leveraged available data for the genetically similar SARS coronavirus to provide specific recommendations for immunological experiments and vaccine design. We developed a web-based platform that regularly monitors new data as the COVID-19 pandemic progresses and updates these recommendations. Moreover, we described the emerging landscape of T cell immune targets by compiling and analyzing SARS-CoV-2-specific data from multiple immunological studies and complemented this with an online platform. Lastly, we reviewed and compared the performance of computational methods that predict T cell targets and have assisted experimental studies on SARS-CoV-2.