UC San Diego

Influenza, despite recently being pushed off to the wayside in the scope of public attention, has placed itself front and center as a recurring potential epidemic. Influenza kills a significant amount of people yearly, and in certain historic cases caused a worldwide epidemic. Previous studies of influenza in the Amaro lab brought light to the function of neuraminidase especially in certain binding sites such as the 450 and 150-loops. The full scale molecular dynamic simulations of mesoscale viral models provided an enormous amount of data that could be exploited to analyze the biological and structural properties of influenza glycoproteins, namely neuraminidase and hemagglutinin. The multiple copies of neuraminidase contained within these whole-virion models allowed for the application of Markov state models to study the dynamics of the neuraminidase 150 and 430-loops. Moreover, our results show that small changes in glycosylation lead to significant differences in the kinetics and morphology of the open and closed conformation of the neuraminidase 150 loop. However, the number of states between the models did not change, which consolidated our current understanding of the 150 -loop dynamics having a clear binary open-closed state. In this thesis, Chapter 1 provides background information about influenza and presents the principles of Markov state model theory. Chapter 2 focuses on the workflow pertinent to Markov State Model building, detailing the process that led to choosing, tweaking and assessing the optimal MSM parameters and settings. Chapter 3 is comprised of the results of the MSM analysis of the neuraminidase 150-loop dynamics. Chapter 4 is an in-depth discussion of the results including future goals and directions.