Dan Larremore, Department of Computer Science, University of Colorado Boulder
Complex Networks, Math, and Malaria: From Evolution to Epidemiology
Progress in the global battle for malaria elimination has flatlined since 2015, with the single-cell P. falciparum parasite killing one child for every minute of the year. Some of this plateau can be attributed to the usual suspects—drug-resistant parasites, insecticide-resilient mosquitos, and counterfeit pharmaceuticals. But fundamentally, malaria persists because parasites prolong infections, evade immune systems, and adapt to individuals and populations. In this talk, I will describe ongoing work to analyze malaria parasites' adaptation strategies. In particular, we will focus on analyzing the parasites' var genes, whose expression allows parasites to evade human immune responses indefinitely. Importantly, due to their complex structure, var genes defy traditional phylogenetic analysis methods and have therefore inspired the development of new and more sophisticated mathematical tools. I will introduce two of those tools, focusing on recent advances in (1) methods for community detection in bipartite networks and (2) Bayes-optimal inference of parasite relatedness. We'll then explore how they can be used to answer open questions in malaria's evolution and genetic epidemiology.