OOS 58-3 - Comparative virology: Challenges in understanding virus-host dynamics and eco-evolutionary feedback mechanisms in virology and virus ecology

Unlike physics and chemistry for which well-defined laws are established, "rules of life" in biology are not as well constructed. In virology and virus ecology, the lack of universal laws is due, in part, to the vastness and diversity of the virosphere. To develop rules of life or laws in virology to which all viruses (or, at least, large groups of viruses) adhere, integration of data across scales, domains of life, and fields of science is required. Indeed, integration of data at the intersections of molecular virology, cell physiology, genetics/genomics, bioinformatics, microbial ecology, biogeochemistry, and computational modeling are minimally required. In this presentation, we introduce a body of ongoing work funded by a U.S. National Science Foundation Biology Integration Institute grant focused on tackling some of the most significant challenges hindering the development of universal laws in virology. Even making comparisons between model virus systems to establish a set of rules that is generalizable across sectors of the virosphere is challenging. Often, the obstacles are due to seemingly simple problems, such as having a common way to accurately determine virus titer in different virus systems. We address some of these major challenges as well as strategies to overcome limitations in methods and confounds stemming from diverse system properties so that valid, quantitative comparisons can be made between disparate virus systems (even those which occupy different domains of life). In addition to standardizing methods for studying virus-host systems, we demonstrate how relationships between variables may be mathematically characterized to yield indices and other measures that may be used to parametrize mathematical models. Such models may prove useful in forecasting outcomes of virus-host interactions, co-evolutionary trajectories, and eco-evolutionary feedback dynamics. This work seeks to advance efforts in characterizing virus-host dynamics so that models with predictive power may be developed. Model validation can be accomplished using directed-evolution experiments and in silico experiments driven by machine learning algorithms. The development of mathematically characterizable relationships between key properties of virus-host systems will be a step forward in the goal to more accurately predict: the potential for species jump events, the emergence of highly virulent virus strains, transmission dynamics, time-course of attenuation, and larger scale impacts of virus-host interactions on ecosystems.