Phylogenetics, Within-Host Evolution, and Who-Infected-Whom

• Activity Number: 493
• Type: Topic Contributed
• Date/Time: Wednesday, August 6, 2014 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Eben Kenah*+
• Companies: University of Florida

Abstract

Many applications of phylogenetics in infectious disease epidemiology use coalescent models and population-level epidemic models. As whole-genome sequences become more readily available, pathogen genetic sequences will become a useful source of information for high-resolution epidemiologic studies of households, schools, and other close-contact groups. A phylogenetic tree linking pathogens sampled from infected people contains information about who-infected-whom, but the relationship between a pathogen phylogeny and who-infected-whom is complicated by within-host evolution. Here, we describe algorithms for finding all possible combinations of who-infected-whom given a phylogenetic tree. We show how likelihood calculation depends on within-host evolutionary models and outline how these algorithms could be incorporated into a Bayesian analysis integrating genetic data with epidemiologic data on person, place, and time. We illustrate our methods with simulations and with an analysis of household data on seasonal and pandemic influenza from Hong Kong. Finally, we discuss how these methods can be extended to semiparametric models of disease transmission.