Denise Kühnert

BEAST 2: A Software Platform for Bayesian Evolutionary Analysis

We present a new open source, extensible and flexible software platform for Bayesian evolutionary analysis called BEAST 2. This software platform is a re-design of the popular BEAST 1 platform to correct structural deficiencies that became evident as the BEAST 1 software evolved. Key among those deficiencies was the lack of post-deployment extensibility. BEAST 2 now has a fully developed package management system that allows third party developers to write additional functionality that can be directly installed to the BEAST 2 analysis platform via a package manager without requiring a new software release of the platform. This package architecture is showcased with a number of recently published new models encompassing birth-death-sampling tree priors, phylodynamics and model averaging for substitution models and site partitioning. A second major improvement is the ability to read/write the entire state of the MCMC chain to/from disk allowing it to be easily shared between multiple instances of the BEAST software. This facilitates checkpointing and better support for multi-processor and high-end computing extensions. Finally, the functionality in new packages can be easily added to the user interface (BEAUti 2) by a simple XML template-based mechanism because BEAST 2 has been re-designed to provide greater integration between the analysis engine and the user interface so that, for example BEAST and BEAUti use exactly the same XML file format.

Birth–death skyline plot reveals temporal changes of epidemic spread in HIV and hepatitis C virus (HCV)

Phylogenetic trees can be used to infer the processes that generated them. Here, we introduce a model, the Bayesian birth–death skyline plot, which explicitly estimates the rate of transmission, recovery, and sampling and thus allows inference of the effective reproductive number directly from genetic data. Our method allows these parameters to vary through time in a piecewise fashion and is implemented within the BEAST2 software framework. The method is a powerful alternative to the existing coalescent skyline plot, providing insight into the differing roles of incidence and prevalence in an epidemic. We apply this method to data from the United Kingdom HIV-1 epidemic and Egyptian hepatitis C virus (HCV) epidemic. The analysis reveals temporal changes of the effective reproductive number that highlight the effect of past public health interventions.

Influenza A Virus Migration and Persistence in North American Wild Birds

Wild birds have been implicated in the emergence of human and livestock influenza. The successful prediction of viral spread and disease emergence, as well as formulation of preparedness plans have been hampered by a critical lack of knowledge of viral movements between different host populations. The patterns of viral spread and subsequent risk posed by wild bird viruses therefore remain unpredictable. Here we analyze genomic data, including 287 newly sequenced avian influenza A virus (AIV) samples isolated over a 34-year period of continuous systematic surveillance of North American migratory birds. We use a Bayesian statistical framework to test hypotheses of viral migration, population structure and patterns of genetic reassortment. Our results reveal that despite the high prevalence of Charadriiformes infected in Delaware Bay this host population does not appear to significantly contribute to the North American AIV diversity sampled in Anseriformes. In contrast, influenza viruses sampled from Anseriformes in Alberta are representative of the AIV diversity circulating in North American Anseriformes. While AIV may be restricted to specific migratory flyways over short time frames, our large-scale analysis showed that the long-term persistence of AIV was independent of bird flyways with migration between populations throughout North America. Analysis of long-term surveillance data provides vital insights to develop appropriately informed predictive models critical for pandemic preparedness and livestock protection.

The contrasting phylodynamics of human influenza B viruses

A complex interplay of viral, host, and ecological factors shapes the spatio-temporal incidence and evolution of human influenza viruses. Although considerable attention has been paid to influenza A viruses, a lack of equivalent data means that an integrated evolutionary and epidemiological framework has until now not been available for influenza B viruses, despite their significant disease burden. Through the analysis of over 900 full genomes from an epidemiological collection of more than 26,000 strains from Australia and New Zealand, we reveal fundamental differences in the phylodynamics of the two co-circulating lineages of influenza B virus (Victoria and Yamagata), showing that their individual dynamics are determined by a complex relationship between virus transmission, age of infection, and receptor binding preference. In sum, this work identifies new factors that are important determinants of influenza B evolution and epidemiology. DOI: http://dx.doi.org/10.7554/eLife.05055.001

Insights into the Early Epidemic Spread of Ebola in Sierra Leone Provided by Viral Sequence Data

Background and Methodology: The current Ebola virus epidemic in West Africa has been spreading at least since December 2013. The first confirmed case of Ebola virus in Sierra Leone was identified on May 25. Based on viral genetic sequencing data from 72 individuals in Sierra Leone collected between the end of May and mid June, we utilize a range of phylodynamic methods to estimate the basic reproductive number (R0). We additionally estimate the expected lengths of the incubation and infectious periods of the virus. Finally, we use phylogenetic trees to examine the role played by population structure in the epidemic. Results: The median estimates of R0 based on sequencing data alone range between 1.65-2.18, with the most plausible model yielding a median R0 of 2.18 (95% HPD 1.24-3.55). Importantly, our results indicate that, at least until mid June, relief efforts in Sierra Leone were ineffective at lowering the effective reproductive number of the virus. We estimate the expected length of the infectious period to be 2.58 days (median; 95% HPD 1.24-6.98). The dataset appears to be too small in order to estimate the incubation period with high certainty (median expected incubation period 4.92 days; 95% HPD 2.11-23.20). While our estimates of the duration of infection tend to be smaller than previously reported, phylodynamic analyses support a previous estimate that 70% of cases were observed and included in the present dataset. The dataset is too small to show a particular population structure with high significance, however our preliminary analyses suggest that half the population is spreading the virus with an R0 well above 2, while the other half of the population is spreading with an R0 below 1. Conclusions: Overall we show that sequencing data can robustly infer key epidemiological parameters. Such estimates inform public health officials and help to coordinate effective public health efforts. Thus having more sequencing data available for the ongoing Ebola virus epidemic and at the start of new outbreaks will foster a quick understanding of the dynamics of the pathogen.

Phylogenetic and epidemic modeling of rapidly evolving infectious diseases.

Abstract Epidemic modeling of infectious diseases has a long history in both theoretical and empirical research. However the recent explosion of genetic data has revealed the rapid rate of evolution that many populations of infectious agents undergo and has underscored the need to consider both evolutionary and ecological processes on the same time scale. Mathematical epidemiology has applied dynamical models to study infectious epidemics, but these models have tended not to exploit – or take into account – evolutionary changes and their effect on the ecological processes and population dynamics of the infectious agent. On the other hand, statistical phylogenetics has increasingly been applied to the study of infectious agents. This approach is based on phylogenetics, molecular clocks, genealogy-based population genetics and phylogeography. Bayesian Markov chain Monte Carlo and related computational tools have been the primary source of advances in these statistical phylogenetic approaches. Recently the first tentative steps have been taken to reconcile these two theoretical approaches. We survey the Bayesian phylogenetic approach to epidemic modeling of infection diseases and describe the contrasts it provides to mathematical epidemiology as well as emphasize the significance of the future unification of these two fields.

Efficient Bayesian inference under the structured coalescent

Motivation: Population structure significantly affects evolutionary dynamics. Such structure may be due to spatial segregation, but may also reflect any other gene-flow-limiting aspect of a model. In combination with the structured coalescent, this fact can be used to inform phylogenetic tree reconstruction, as well as to infer parameters such as migration rates and subpopulation sizes from annotated sequence data. However, conducting Bayesian inference under the structured coalescent is impeded by the difficulty of constructing Markov Chain Monte Carlo (MCMC) sampling algorithms (samplers) capable of efficiently exploring the state space. Results: In this article, we present a new MCMC sampler capable of sampling from posterior distributions over structured trees: timed phylogenetic trees in which lineages are associated with the distinct subpopulation in which they lie. The sampler includes a set of MCMC proposal functions that offer significant mixing improvements over a previously published method. Furthermore, its implementation as a BEAST 2 package ensures maximum flexibility with respect to model and prior specification. We demonstrate the usefulness of this new sampler by using it to infer migration rates and effective population sizes of H3N2 influenza between New Zealand, New York and Hong Kong from publicly available hemagglutinin (HA) gene sequences under the structured coalescent. Availability and implementation: The sampler has been implemented as a publicly available BEAST 2 package that is distributed under version 3 of the GNU General Public License at http://compevol.github.io/MultiTypeTree. Contact: tgvaughan@gmail.com Supplementary information: Supplementary data are available at Bioinformatics online.

Origin of modern syphilis and emergence of a pandemic Treponema pallidum cluster

The abrupt onslaught of the syphilis pandemic that started in the late fifteenth century established this devastating infectious disease as one of the most feared in human history1. Surprisingly, despite the availability of effective antibiotic treatment since the mid-twentieth century, this bacterial infection, which is caused by Treponema pallidum subsp. pallidum (TPA), has been re-emerging globally in the last few decades with an estimated 10.6 million cases in 2008 (ref. 2). Although resistance to penicillin has not yet been identified, an increasing number of strains fail to respond to the second-line antibiotic azithromycin3. Little is known about the genetic patterns in current infections or the evolutionary origins of the disease due to the low quantities of treponemal DNA in clinical samples and difficulties in cultivating the pathogen4. Here, we used DNA capture and whole-genome sequencing to successfully interrogate genome-wide variation from syphilis patient specimens, combined with laboratory samples of TPA and two other subspecies. Phylogenetic comparisons based on the sequenced genomes indicate that the TPA strains examined share a common ancestor after the fifteenth century, within the early modern era. Moreover, most contemporary strains are azithromycin-resistant and are members of a globally dominant cluster, named here as SS14-Ω. The cluster diversified from a common ancestor in the mid-twentieth century subsequent to the discovery of antibiotics. Its recent phylogenetic divergence and global presence point to the emergence of a pandemic strain cluster.

Phylodynamics of Enterovirus A71-Associated Hand, Foot, and Mouth Disease in Viet Nam

ABSTRACT Enterovirus A71 (EV-A71) is a major cause of hand, foot, and mouth disease (HFMD) and is particularly prevalent in parts of Southeast Asia, affecting thousands of children and infants each year. Revealing the evolutionary and epidemiological dynamics of EV-A71 through time and space is central to understanding its outbreak potential. We generated the full genome sequences of 200 EV-A71 strains sampled from various locations in Viet Nam between 2011 and 2013 and used these sequence data to determine the evolutionary history and phylodynamics of EV-A71 in Viet Nam, providing estimates of the effective reproduction number (Re) of the infection through time. In addition, we described the phylogeography of EV-A71 throughout Southeast Asia, documenting patterns of viral gene flow. Accordingly, our analysis reveals that a rapid genogroup switch from C4 to B5 likely took place during 2012 in Viet Nam. We show that the Re of subgenogroup C4 decreased during the time frame of sampling, whereas that of B5 increased and remained >1 at the end of 2013, corresponding to a rise in B5 prevalence. Our study reveals that the subgenogroup B5 virus that emerged into Viet Nam is closely related to variants that were responsible for large epidemics in Malaysia and Taiwan and therefore extends our knowledge regarding its associated area of endemicity. Subgenogroup B5 evidently has the potential to cause more widespread outbreaks across Southeast Asia. IMPORTANCE EV-A71 is one of many viruses that cause HFMD, a common syndrome that largely affects infants and children. HFMD usually causes only mild illness with no long-term consequences. Occasionally, however, severe infection may arise, especially in very young children, causing neurological complications and even death. EV-A71 is highly contagious and is associated with the most severe HFMD cases, with large and frequent epidemics of the virus recorded worldwide. Although major advances have been made in the development of a potential EV-A71 vaccine, there is no current prevention and little is known about the patterns and dynamics of EV-A71 spread. In this study, we utilize full-length genome sequence data obtained from HFMD patients in Viet Nam, a geographical region where the disease has been endemic since 2003, to characterize the phylodynamics of this important emerging virus.

Phylodynamics with Migration: A Computational Framework to Quantify Population Structure from Genomic Data

When viruses spread, outbreaks can be spawned in previously unaffected regions. Depending on the time and mode of introduction, each regional outbreak can have its own epidemic dynamics. The migration and phylodynamic processes are often intertwined and need to be taken into account when analyzing temporally and spatially structured virus data. In this article, we present a fully probabilistic approach for the joint reconstruction of phylodynamic history in structured populations (such as geographic structure) based on a multitype birth–death process. This approach can be used to quantify the spread of a pathogen in a structured population. Changes in epidemic dynamics through time within subpopulations are incorporated through piecewise constant changes in transmission parameters. We analyze a global human influenza H3N2 virus data set from a geographically structured host population to demonstrate how seasonal dynamics can be inferred simultaneously with the phylogeny and migration process. Our results suggest that the main migration path among the northern, tropical, and southern region represented in the sample analyzed here is the one leading from the tropics to the northern region. Furthermore, the time-dependent transmission dynamics between and within two HIV risk groups, heterosexuals and injecting drug users, in the Latvian HIV epidemic are investigated. Our analyses confirm that the Latvian HIV epidemic peaking around 2001 was mainly driven by the injecting drug user risk group.