Inferring the Source of Transmission with Phylogenetic Data

Identifying the source of transmission using pathogen genetic data is complicated by numerous biological, immunological, and behavioral factors. A large source of error arises when there is incomplete or sparse sampling of cases. Unsampled cases may act as either a common source of infection or as an intermediary in a transmission chain for hosts infected with genetically similar pathogens. It is difficult to quantify the probability of common source or intermediate transmission events, which has made it difficult to develop statistical tests to either confirm or deny putative transmission pairs with genetic data. We present a method to incorporate additional information about an infectious disease epidemic, such as incidence and prevalence of infection over time, to inform estimates of the probability that one sampled host is the direct source of infection of another host in a pathogen gene genealogy. These methods enable forensic applications, such as source-case attribution, for infectious disease epidemics with incomplete sampling, which is usually the case for high-morbidity community-acquired pathogens like HIV, Influenza and Dengue virus. These methods also enable epidemiological applications such as the identification of factors that increase the risk of transmission. We demonstrate these methods in the context of the HIV epidemic in Detroit, Michigan, and we evaluate the suitability of current sequence databases for forensic and epidemiological investigations. We find that currently available sequences collected for drug resistance testing of HIV are unlikely to be useful in most forensic investigations, but are useful for identifying transmission risk factors.     

Regression-Based Ranking of Pathogen Strains with Respect to Their Contribution to Natural Epidemics

Genetic variation in pathogen populations may be an important factor driving heterogeneity in disease dynamics within their host populations. However, to date, we understand poorly how genetic diversity in diseases impact on epidemiological dynamics because data and tools required to answer this questions are lacking. Here, we combine pathogen genetic data with epidemiological monitoring of disease progression, and introduce a statistical exploratory method to investigate differences among pathogen strains in their performance in the field. The method exploits epidemiological data providing a measure of disease progress in time and space, and genetic data indicating the relative spatial patterns of the sampled pathogen strains. Applying this method allows to assign ranks to the pathogen strains with respect to their contributions to natural epidemics and to assess the significance of the ranking. This method was first tested on simulated data, including data obtained from an original, stochastic, multi-strain epidemic model. It was then applied to epidemiological and genetic data collected during one natural epidemic of powdery mildew occurring in its wild host population. Based on the simulation study, we conclude that the method can achieve its aim of ranking pathogen strains if the sampling effort is sufficient. For powdery mildew data, the method indicated that one of the sampled strains tends to have a higher fitness than the four other sampled strains, highlighting the importance of strain diversity for disease dynamics. Our approach allowing the comparison of pathogen strains in natural epidemic is complementary to the classical practice of using experimental infections in controlled conditions to estimate fitness of different pathogen strains. Our statistical tool, implemented in the R package StrainRanking, is mainly based on regression and does not rely on mechanistic assumptions on the pathogen dynamics. Thus, the method can be applied to a wide range of pathogens.     

A computationally tractable birth-death model that combines phylogenetic and epidemiological data

Inferring the dynamics of pathogen transmission during an outbreak is an important problem in infectious disease epidemiology. In mathematical epidemiology, estimates are often informed by time series of confirmed cases, while in phylodynamics genetic sequences of the pathogen, sampled through time, are the primary data source. Each type of data provides different, and potentially complementary, insight. Recent studies have recognised that combining data sources can improve estimates of the transmission rate and the number of infected individuals. However, inference methods are typically highly specialised and field-specific and are either computationally prohibitive or require intensive simulation, limiting their real-time utility. We present a novel birth-death phylogenetic model and derive a tractable analytic approximation of its likelihood, the computational complexity of which is linear in the size of the dataset. This approach combines epidemiological and phylodynamic data to produce estimates of key parameters of transmission dynamics and the unobserved prevalence. Using simulated data, we show (a) that the approximation agrees well with existing methods, (b) validate the claim of linear complexity and (c) explore robustness to model misspecification. This approximation facilitates inference on large datasets, which is increasingly important as large genomic sequence datasets become commonplace.     

Unravelling transmission trees of infectious diseases by combining genetic and epidemiological data

Knowledge on the transmission tree of an epidemic can provide valuable insights into disease dynamics. The transmission tree can be reconstructed by analysing either detailed epidemiological data (e.g. contact tracing) or, if sufficient genetic diversity accumulates over the course of the epidemic, genetic data of the pathogen. We present a likelihood-based framework to integrate these two data types, estimating probabilities of infection by taking weighted averages over the set of possible transmission trees. We test the approach by applying it to temporal, geographical and genetic data on the 241 poultry farms infected in an epidemic of avian influenza A (H7N7) in The Netherlands in 2003. We show that the combined approach estimates the transmission tree with higher correctness and resolution than analyses based on genetic or epidemiological data alone. Furthermore, the estimated tree reveals the relative infectiousness of farms of different types and sizes.     

Within-Host Bacterial Diversity Hinders Accurate Reconstruction of Transmission Networks from Genomic Distance Data

The prospect of using whole genome sequence data to investigate bacterial disease outbreaks has been keenly anticipated in many quarters, and the large-scale collection and sequencing of isolates from cases is becoming increasingly feasible. While sequence data can provide many important insights into disease spread and pathogen adaptation, it remains unclear how successfully they may be used to estimate individual routes of transmission. Several studies have attempted to reconstruct transmission routes using genomic data; however, these have typically relied upon restrictive assumptions, such as a shared topology of the phylogenetic tree and a lack of within-host diversity. In this study, we investigated the potential for bacterial genomic data to inform transmission network reconstruction. We used simulation models to investigate the origins, persistence and onward transmission of genetic diversity, and examined the impact of such diversity on our estimation of the epidemiological relationship between carriers. We used a flexible distance-based metric to provide a weighted transmission network, and used receiver-operating characteristic (ROC) curves and network entropy to assess the accuracy and uncertainty of the inferred structure. Our results suggest that sequencing a single isolate from each case is inadequate in the presence of within-host diversity, and is likely to result in misleading interpretations of transmission dynamics – under many plausible conditions, this may be little better than selecting transmission links at random. Sampling more frequently improves accuracy, but much uncertainty remains, even if all genotypes are observed. While it is possible to discriminate between clusters of carriers, individual transmission routes cannot be resolved by sequence data alone. Our study demonstrates that bacterial genomic distance data alone provide only limited information on person-to-person transmission dynamics.     

Finding Evidence for Local Transmission of Contagious Disease in Molecular Epidemiological Datasets

Surveillance systems of contagious diseases record information on cases to monitor incidence of disease and to evaluate effectiveness of interventions. These systems focus on a well-defined population; a key question is whether observed cases are infected through local transmission within the population or whether cases are the result of importation of infection into the population. Local spread of infection calls for different intervention measures than importation of infection. Besides standardized information on time of symptom onset and location of cases, pathogen genotyping or sequencing offers essential information to address this question. Here we introduce a method that takes full advantage of both the genetic and epidemiological data to distinguish local transmission from importation of infection, by comparing inter-case distances in temporal, spatial and genetic data. Cases that are part of a local transmission chain will have shorter distances between their geographical locations, shorter durations between their times of symptom onset and shorter genetic distances between their pathogen sequences as compared to cases that are due to importation. In contrast to generic clustering algorithms, the proposed method explicitly accounts for the fact that during local transmission of a contagious disease the cases are caused by other cases. No pathogen-specific assumptions are needed due to the use of ordinal distances, which allow for direct comparison between the disparate data types. Using simulations, we test the performance of the method in identifying local transmission of disease in large datasets, and assess how sensitivity and specificity change with varying size of local transmission chains and varying overall disease incidence.     

Inference of Epidemiological Dynamics Based on Simulated Phylogenies Using Birth-Death and Coalescent Models

Quantifying epidemiological dynamics is crucial for understanding and forecasting the spread of an epidemic. The coalescent and the birth-death model are used interchangeably to infer epidemiological parameters from the genealogical relationships of the pathogen population under study, which in turn are inferred from the pathogen genetic sequencing data. To compare the performance of these widely applied models, we performed a simulation study. We simulated phylogenetic trees under the constant rate birth-death model and the coalescent model with a deterministic exponentially growing infected population. For each tree, we re-estimated the epidemiological parameters using both a birth-death and a coalescent based method, implemented as an MCMC procedure in BEAST v2.0. In our analyses that estimate the growth rate of an epidemic based on simulated birth-death trees, the point estimates such as the maximum a posteriori/maximum likelihood estimates are not very different. However, the estimates of uncertainty are very different. The birth-death model had a higher coverage than the coalescent model, i.e. contained the true value in the highest posterior density (HPD) interval more often (2–13% vs. 31–75% error). The coverage of the coalescent decreases with decreasing basic reproductive ratio and increasing sampling probability of infecteds. We hypothesize that the biases in the coalescent are due to the assumption of deterministic rather than stochastic population size changes. Both methods performed reasonably well when analyzing trees simulated under the coalescent. The methods can also identify other key epidemiological parameters as long as one of the parameters is fixed to its true value. In summary, when using genetic data to estimate epidemic dynamics, our results suggest that the birth-death method will be less sensitive to population fluctuations of early outbreaks than the coalescent method that assumes a deterministic exponentially growing infected population.     

Bayesian Reconstruction of Disease Outbreaks by Combining Epidemiologic and Genomic Data

Recent years have seen progress in the development of statistically rigorous frameworks to infer outbreak transmission trees (“who infected whom”) from epidemiological and genetic data. Making use of pathogen genome sequences in such analyses remains a challenge, however, with a variety of heuristic approaches having been explored to date. We introduce a statistical method exploiting both pathogen sequences and collection dates to unravel the dynamics of densely sampled outbreaks. Our approach identifies likely transmission events and infers dates of infections, unobserved cases and separate introductions of the disease. It also proves useful for inferring numbers of secondary infections and identifying heterogeneous infectivity and super-spreaders. After testing our approach using simulations, we illustrate the method with the analysis of the beginning of the 2003 Singaporean outbreak of Severe Acute Respiratory Syndrome (SARS), providing new insights into the early stage of this epidemic. Our approach is the first tool for disease outbreak reconstruction from genetic data widely available as free software, the R package outbreaker. It is applicable to various densely sampled epidemics, and improves previous approaches by detecting unobserved and imported cases, as well as allowing multiple introductions of the pathogen. Because of its generality, we believe this method will become a tool of choice for the analysis of densely sampled disease outbreaks, and will form a rigorous framework for subsequent methodological developments.     

Integrating genetic and epidemiological data to determine transmission pathways of foot-and-mouth disease virus

Estimating detailed transmission trees that reflect the relationships between infected individuals or populations during a disease outbreak often provides valuable insights into both the nature of disease transmission and the overall dynamics of the underlying epidemiological process. These trees may be based on epidemiological data that relate to the timing of infection and infectiousness, or genetic data that show the genetic relatedness of pathogens isolated from infected individuals. Genetic data are becoming increasingly important in the estimation of transmission trees of viral pathogens due to their inherently high mutation rate. Here, we propose a maximum-likelihood approach that allows epidemiological and genetic data to be combined within the same analysis to infer probable transmission trees. We apply this approach to data from 20 farms infected during the 2001 UK foot-and-mouth disease outbreak, using complete viral genome sequences from each infected farm and information on when farms were first estimated to have developed clinical disease and when livestock on these farms were culled. Incorporating known infection links due to animal movement prior to imposition of the national movement ban results in the reduction of the number of trees from 41472 that are consistent with the genetic data to 1728, of which just 4 represent more than 95% of the total likelihood calculated using a model that accounts for the epidemiological data. These trees differ in several ways from those constructed prior to the availability of genetic data.     

SARS transmission pattern in Singapore reassessed by viral sequence variation analysis

Background

Epidemiological investigations of infectious disease are mainly dependent on indirect contact information and only occasionally assisted by characterization of pathogen sequence variation from clinical isolates. Direct sequence analysis of the pathogen, particularly at a population level, is generally thought to be too cumbersome, technically difficult, and expensive. We present here a novel application of mass spectrometry (MS)–based technology in characterizing viral sequence variations that overcomes these problems, and we apply it retrospectively to the severe acute respiratory syndrome (SARS) outbreak in Singapore.

Methods and Findings

The success rate of the MS-based analysis for detecting SARS coronavirus (SARS-CoV) sequence variations was determined to be 95% with 75 copies of viral RNA per reaction, which is sufficient to directly analyze both clinical and cultured samples. Analysis of 13 SARS-CoV isolates from the different stages of the Singapore outbreak identified nine sequence variations that could define the molecular relationship between them and pointed to a new, previously unidentified, primary route of introduction of SARS-CoV into the Singapore population. Our direct determination of viral sequence variation from a clinical sample also clarified an unresolved epidemiological link regarding the acquisition of SARS in a German patient. We were also able to detect heterogeneous viral sequences in primary lung tissues, suggesting a possible coevolution of quasispecies of virus within a single host.

Conclusion

This study has further demonstrated the importance of improving clinical and epidemiological studies of pathogen transmission through the use of genetic analysis and has revealed the MS-based analysis to be a sensitive and accurate method for characterizing SARS-CoV genetic variations in clinical samples. We suggest that this approach should be used routinely during outbreaks of a wide variety of agents, in order to allow the most effective control.     

Elucidating relationships between P.falciparum prevalence and measures of genetic diversity with a combined genetic-epidemiological model of malaria

There is an abundance of malaria genetic data being collected from the field, yet using these data to understand the drivers of regional epidemiology remains a challenge. A key issue is the lack of models that relate parasite genetic diversity to epidemiological parameters. Classical models in population genetics characterize changes in genetic diversity in relation to demographic parameters, but fail to account for the unique features of the malaria life cycle. In contrast, epidemiological models, such as the Ross-Macdonald model, capture malaria transmission dynamics but do not consider genetics. Here, we have developed an integrated model encompassing both parasite evolution and regional epidemiology. We achieve this by combining the Ross-Macdonald model with an intra-host continuous-time Moran model, thus explicitly representing the evolution of individual parasite genomes in a traditional epidemiological framework. Implemented as a stochastic simulation, we use the model to explore relationships between measures of parasite genetic diversity and parasite prevalence, a widely-used metric of transmission intensity. First, we explore how varying parasite prevalence influences genetic diversity at equilibrium. We find that multiple genetic diversity statistics are correlated with prevalence, but the strength of the relationships depends on whether variation in prevalence is driven by host- or vector-related factors. Next, we assess the responsiveness of a variety of statistics to malaria control interventions, finding that those related to mixed infections respond quickly (∼months) whereas other statistics, such as nucleotide diversity, may take decades to respond. These findings provide insights into the opportunities and challenges associated with using genetic data to monitor malaria epidemiology.     

A Bayesian Inference Framework to Reconstruct Transmission Trees Using Epidemiological and Genetic Data

The accurate identification of the route of transmission taken by an infectious agent through a host population is critical to understanding its epidemiology and informing measures for its control. However, reconstruction of transmission routes during an epidemic is often an underdetermined problem: data about the location and timings of infections can be incomplete, inaccurate, and compatible with a large number of different transmission scenarios. For fast-evolving pathogens like RNA viruses, inference can be strengthened by using genetic data, nowadays easily and affordably generated. However, significant statistical challenges remain to be overcome in the full integration of these different data types if transmission trees are to be reliably estimated. We present here a framework leading to a bayesian inference scheme that combines genetic and epidemiological data, able to reconstruct most likely transmission patterns and infection dates. After testing our approach with simulated data, we apply the method to two UK epidemics of Foot-and-Mouth Disease Virus (FMDV): the 2007 outbreak, and a subset of the large 2001 epidemic. In the first case, we are able to confirm the role of a specific premise as the link between the two phases of the epidemics, while transmissions more densely clustered in space and time remain harder to resolve. When we consider data collected from the 2001 epidemic during a time of national emergency, our inference scheme robustly infers transmission chains, and uncovers the presence of undetected premises, thus providing a useful tool for epidemiological studies in real time. The generation of genetic data is becoming routine in epidemiological investigations, but the development of analytical tools maximizing the value of these data remains a priority. Our method, while applied here in the context of FMDV, is general and with slight modification can be used in any situation where both spatiotemporal and genetic data are available.     

猪链球菌分子流行病学研究方法

猪链球菌（Streptococcus suis）是猪重要的细菌性病原,它可以导致猪的脑膜炎、败血症和关节炎等症状,给养猪业带来严重经济损失;同时该菌还可感染人,是一种人畜共患病原菌。应用分子流行病学方法,阐明猪链球菌病的流行病学特征,明确其毒力分型、时空分布、传播途径、传染源,确定传播的遗传决定因素等,将有助于猪链球菌病的防控。目前常用的分子流行病学方法主要有多位点序列分型、脉冲场凝胶电泳、全基因组测序和基于PCR的方法等。本文介绍了上述方法的原理以及在猪链球菌流行病学中的应用,并分析这几种方法的优缺点,从而为更好地揭示猪链球菌流行病学特征、制定猪链球菌病的防控策略提供参考。     

BDM1, a phosducin-like gene of Fusarium graminearum, is involved in virulence during infection of wheat and maize

Fusarium graminearum is a common pathogen of wheat and maize throughout the world. Despite recent advances in the elucidation of the genetic basis of virulence, significant gaps in the regulatory network underlying pathogenesis remain to be filled. In particular, little is known at the molecular level about the overlap among mechanisms of pathogenicity on maize and wheat. G-protein signalling has been implicated in pathogenesis in F. graminearum, although the underlying mechanisms are not fully understood. In this study, we investigated the involvement of a putative phosducin-like gene (BDM1) in growth, development and pathogenesis in F. graminearum. Targeted deletion of BDM1 revealed roles in sexual and asexual sporulation, germ tube development, hyphal branching and mycelial morphology. During pathogenesis, BDM1 is required for wild-type levels of colonization of maize silk tissue and stalks, but is dispensable for the colonization of kernels. The deletion of BDM1 also reduced the virulence of F. graminearum during the infection of wheat seedlings and heads, resulting in a significant reduction in fungal biomass and a delayed spread of visual symptom expression (i.e. bleaching in heads). Furthermore, BDM1 is required for wild-type levels of deoxynivalenol biosynthesis during the infection of wheat heads and maize silks. In summation, BDM1 is one of the few genes characterized to date in F. graminearum involved in virulence during infection of both maize and wheat. Thus, the functional characterization of BDM1 has established a new regulatory link between pathogenesis in maize and wheat, and provides a genetic resource through which the regulatory networks underlying virulence in F. graminearum can be further elucidated.     

'SEEDY' (Simulation of Evolutionary and Epidemiological Dynamics): An R Package to Follow Accumulation of Within-Host Mutation in Pathogens

Genome sequencing is an increasingly common component of infectious disease outbreak investigations. However, the relationship between pathogen transmission and observed genetic data is complex, and dependent on several uncertain factors. As such, simulation of pathogen dynamics is an important tool for interpreting observed genomic data in an infectious disease outbreak setting, in order to test hypotheses and to explore the range of outcomes consistent with a given set of parameters. We introduce ‘seedy’, an R package for the simulation of evolutionary and epidemiological dynamics (http://cran.r-project.org/web/packages/seedy/). Our software implements stochastic models for the accumulation of mutations within hosts, as well as individual-level disease transmission. By allowing variables such as the transmission bottleneck size, within-host effective population size and population mixing rates to be specified by the user, our package offers a flexible framework to investigate evolutionary dynamics during disease outbreaks. Furthermore, our software provides theoretical pairwise genetic distance distributions to provide a likelihood of person-to-person transmission based on genomic observations, and using this framework, implements transmission route assessment for genomic data collected during an outbreak. Our open source software provides an accessible platform for users to explore pathogen evolution and outbreak dynamics via simulation, and offers tools to assess observed genomic data in this context.     

The structural identifiability and parameter estimation of a multispecies model for the transmission of mastitis in dairy cows.

A structural identifiability analysis is performed on a mathematical model for the coupled transmission of two classes of pathogen. The pathogens, classified as major and minor, are aetiological agents of mastitis in dairy cows that interact directly and via the immunological reaction in their hosts. Parameter estimates are available from experimental data for all but four of the parameters in the model. Data from a longitudinal study of infection are used to estimate these unknown parameters. A novel approach and application of structural identifiability analysis is combined in this paper with the estimation of cross-protection parameters using epidemiological data.     

A new phylodynamic model of Mycobacterium bovis transmission in a multi-host system uncovers the role of the unobserved reservoir

Multi-host pathogens are particularly difficult to control, especially when at least one of the hosts acts as a hidden reservoir. Deep sequencing of densely sampled pathogens has the potential to transform this understanding, but requires analytical approaches that jointly consider epidemiological and genetic data to best address this problem. While there has been considerable success in analyses of single species systems, the hidden reservoir problem is relatively under-studied. A well-known exemplar of this problem is bovine Tuberculosis, a disease found in British and Irish cattle caused by Mycobacterium bovis, where the Eurasian badger has long been believed to act as a reservoir but remains of poorly quantified importance except in very specific locations. As a result, the effort that should be directed at controlling disease in badgers is unclear. Here, we analyse densely collected epidemiological and genetic data from a cattle population but do not explicitly consider any data from badgers. We use a simulation modelling approach to show that, in our system, a model that exploits available cattle demographic and herd-to-herd movement data, but only considers the ability of a hidden reservoir to generate pathogen diversity, can be used to choose between different epidemiological scenarios. In our analysis, a model where the reservoir does not generate any diversity but contributes to new infections at a local farm scale are significantly preferred over models which generate diversity and/or spread disease at broader spatial scales. While we cannot directly attribute the role of the reservoir to badgers based on this analysis alone, the result supports the hypothesis that under current cattle control regimes, infected cattle alone cannot sustain M. bovis circulation. Given the observed close phylogenetic relationship for the bacteria taken from cattle and badgers sampled near to each other, the most parsimonious hypothesis is that the reservoir is the infected badger population. More broadly, our approach demonstrates that carefully constructed bespoke models can exploit the combination of genetic and epidemiological data to overcome issues of extreme data bias, and uncover important general characteristics of transmission in multi-host pathogen systems.     

Transmission of Equine Influenza Virus during an Outbreak Is Characterized by Frequent Mixed Infections and Loose Transmission Bottlenecks

The ability of influenza A viruses (IAVs) to cross species barriers and evade host immunity is a major public health concern. Studies on the phylodynamics of IAVs across different scales – from the individual to the population – are essential for devising effective measures to predict, prevent or contain influenza emergence. Understanding how IAVs spread and evolve during outbreaks is critical for the management of epidemics. Reconstructing the transmission network during a single outbreak by sampling viral genetic data in time and space can generate insights about these processes. Here, we obtained intra-host viral sequence data from horses infected with equine influenza virus (EIV) to reconstruct the spread of EIV during a large outbreak. To this end, we analyzed within-host viral populations from sequences covering 90% of the infected yards. By combining gene sequence analyses with epidemiological data, we inferred a plausible transmission network, in turn enabling the comparison of transmission patterns during the course of the outbreak and revealing important epidemiological features that were not apparent using either approach alone. The EIV populations displayed high levels of genetic diversity, and in many cases we observed distinct viral populations containing a dominant variant and a number of related minor variants that were transmitted between infectious horses. In addition, we found evidence of frequent mixed infections and loose transmission bottlenecks in these naturally occurring populations. These frequent mixed infections likely influence the size of epidemics.     

BAsE-Seq: a method for obtaining long viral haplotypes from short sequence reads

Epidemiologists aim to inform the design of public health interventions with evidence on the evolution, emergence and spread of infectious diseases. Sequencing of pathogen genomes, together with date, location, clinical manifestation and other relevant data about sample origins, can contribute to describing nearly every aspect of transmission dynamics, including local transmission and global spread. The analyses of these data have implications for all levels of clinical and public health practice, from institutional infection control to policies for surveillance, prevention and treatment. This review highlights the range of epidemiological questions that can be addressed from the combination of genome sequence and traditional ‘line lists’ (tables of epidemiological data where each line includes demographic and clinical features of infected individuals). We identify opportunities for these data to inform interventions that reduce disease incidence and prevalence. By considering current limitations of, and challenges to, interpreting these data, we aim to outline a research agenda to accelerate the genomics-driven transformation in public health microbiology.     

Two optimal mutation rates in obligate pathogens subject to deleterious mutation

Pathogen species with high mutation rates are likely to accumulate deleterious mutations that reduce their reproductive potential within the host. By altering the within-host growth rate of the pathogen, the deleterious mutation load has the potential to affect epidemiological properties such as prevalence, mean pathogen load, and the mean duration of infections. Here, I examine an epidemiological model that allows for multiple segregating mutations that affect within-host replication efficiency. The model demonstrates a complex range of outcomes depending on pathogen mutation rate, including two distinct, widely separated mutation rates associated with high pathogen prevalence. The low mutation rate prevalence peak is associated with small amounts of genetic diversity within the pathogen population, relatively stable prevalence and infection dynamics, and genetic variation partitioned between hosts. The high mutation rate peak is characterized by considerable genetic diversity both within and between hosts, relatively frequent invasions by more virulent types, and is qualitatively similar to an RNA virus quasispecies. The two prevalence peaks are separated by a valley where natural selection favors evolution toward the optimal within-host state, which is associated with high virulence and relatively rapid host mortality. Both chronic and acute infections are examined using stochastic forward simulations.