In silico modeling of subcellular infection by diverse families of RNA virus

PROJECT SUMMARY/ABSTRACT RNA viruses are the leading source of existing and emerging pathogens. Many species overtake host cells with just a dozen or so viral components, making subcellular infections tractable for mathematical modeling and analysis. Such models have the potential to identify fragilities in a viral life cycle, examine differences in susceptibility among humans, and serve as templates for reconfiguration in response to novel outbreaks. However, it has not been clear how to build such models in a scalable way, and thus fewer than ten have been developed among the several hundred RNA viruses that infect humans. A new modeling strategy was recently proposed, which starts with a common mass-action topology that is then customized to different virus families by parameter inference. The generic approach lumps together biochemical processes that are specific to different virus families, and it is unclear whether these distinctions are needed to create models that are broadly predictive. The objective of this application is to evaluate the relative merits of generic and familyspecific approaches for modeling subcellular infection by RNA viruses. We focus on coxsackievirus B3 and dengue virus as two species of RNA virus from different families (Picornaviridae and Flaviviridae) for which generic and family-specific models are available or immediately feasible. The overarching hypothesis is that RNA viruses are similarly organized around modules for entry, replication, and other core processes, but the modules fundamentally differ by virus family. The specific aims are to: 1) Compare lumped viral entry to family-specific modules; 2) Refactor the viral replication module; and 3) Add antiviral conduits between modules. Our approach leverages deep transcriptome profiles from several thousand single cells and several hundred relevant human organs; it also invokes a new abstraction (the phase-field crystal model) for a key intermediate of RNA viruses. Computational and experimental tests will be performed using temperature as a system-wide perturbation of biochemical rate parameters at surface (33°C), body (37°C), and febrile (40°C) temperatures. This multi-pronged assessment across different modules will clarify a best-available path toward building foundational models for all major families of RNA viruses that infect humans.