A model-based clustering method to detect infectious disease transmission outbreaks from sequence variation.

Clustering infections by genetic similarity is a popular technique for identifying potential outbreaks of infectious disease, in part because sequences are now routinely collected for clinical management of many infections. A diverse number of nonparametric clustering methods have been developed for this purpose. These methods are generally intuitive, rapid to compute, and readily scale with large data sets. However, we have found that nonparametric clustering methods can be biased towards identifying clusters of diagnosis-where individuals are sampled sooner post-infection-rather than the clusters of rapid transmission that are meant to be potential foci for public health efforts. We develop a fundamentally new approach to genetic clustering based on fitting a Markov-modulated Poisson process (MMPP), which represents the evolution of transmission rates along the tree relating different infections. We evaluated this model-based method alongside five nonparametric clustering methods using both simulated and actual HIV sequence data sets. For simulated clusters of rapid transmission, the MMPP clustering method obtained higher mean sensitivity (85%) and specificity (91%) than the nonparametric methods. When we applied these clustering methods to published sequences from a study of HIV-1 genetic clusters in Seattle, USA, we found that the MMPP method categorized about half (46%) as many individuals to clusters compared to the other methods. Furthermore, the mean internal branch lengths that approximate transmission rates were significantly shorter in clusters extracted using MMPP, but not by other methods. We determined that the computing time for the MMPP method scaled linearly with the size of trees, requiring about 30 seconds for a tree of 1,000 tips and about 20 minutes for 50,000 tips on a single computer. This new approach to genetic clustering has significant implications for the application of pathogen sequence analysis to public health, where it is critical to robustly and accurately identify clusters for the most cost-effective deployment of outbreak management and prevention resources.

Reconstructing contact network parameters from viral phylogenies.

Models of the spread of disease in a population often make the simplifying assumption that the population is homogeneously mixed, or is divided into homogeneously mixed compartments. However, human populations have complex structures formed by social contacts, which can have a significant influence on the rate of epidemic spread. Contact network models capture this structure by explicitly representing each contact which could possibly lead to a transmission. We developed a method based on approximate Bayesian computation (ABC), a likelihood-free inference strategy, for estimating structural parameters of the contact network underlying an observed viral phylogeny. The method combines adaptive sequential Monte Carlo for ABC, Gillespie simulation for propagating epidemics though networks, and a kernel-based tree similarity score. We used the method to fit the Barabási-Albert network model to simulated transmission trees, and also applied it to viral phylogenies estimated from ten published HIV sequence datasets. This model incorporates a feature called preferential attachment (PA), whereby individuals with more existing contacts accumulate new contacts at a higher rate. On simulated data, we found that the strength of PA and the number of infected nodes in the network can often be accurately estimated. On the other hand, the mean degree of the network, as well as the total number of nodes, was not estimable with ABC. We observed sub-linear PA power in all datasets, as well as higher PA power in networks of injection drug users. These results underscore the importance of considering contact structures when performing phylodynamic inference. Our method offers the potential to quantitatively investigate the contact network structure underlying viral epidemics.

The spread of hepatitis C virus genotype 1a in North America: a retrospective phylogenetic study.

BACKGROUND: The timing of the initial spread of hepatitis C virus genotype 1a in North America is controversial. In particular, how and when hepatitis C virus reached extraordinary prevalence in specific demographic groups remains unclear. We quantified, using all available hepatitis C virus sequence data and phylodynamic methods, the timing of the spread of hepatitis C virus genotype 1a in North America. METHODS: We screened 45 316 publicly available sequences of hepatitis C virus genotype 1a for location and genotype, and then did phylogenetic analyses of available North American sequences from five hepatitis C virus genes (E1, E2, NS2, NS4B, NS5B), with an emphasis on including as many sequences with early collection dates as possible. We inferred the historical population dynamics of this epidemic for all five gene regions using Bayesian skyline plots. FINDINGS: Most of the spread of genotype 1a in North America occurred before 1965, and the hepatitis C virus epidemic has undergone relatively little expansion since then. The effective population size of the North American epidemic stabilised around 1960. These results were robust across all five gene regions analysed, although analyses of each gene separately show substantial variation in estimates of the timing of the early exponential growth, ranging roughly from 1940 for NS2, to 1965 for NS4B. INTERPRETATION: The expansion of genotype 1a before 1965 suggests that nosocomial or iatrogenic factors rather than past sporadic behavioural risk (ie, experimentation with injection drug use, unsafe tattooing, high risk sex, travel to high endemic areas) were key contributors to the hepatitis C virus epidemic in North America. Our results might reduce stigmatisation around screening and diagnosis, potentially increasing rates of screening and treatment for hepatitis C virus. FUNDING: The Canadian Institutes of Health Research, Michael Smith Foundation for Health Research, and BC Centre for Excellence in HIV/AIDS.

Differential evolution of a CXCR4-using HIV-1 strain in CCR5wt/wt and CCR5∆32/∆32 hosts revealed by longitudinal deep sequencing and phylogenetic reconstruction.

Rare individuals homozygous for a naturally-occurring 32 base pair deletion in the CCR5 gene (CCR5∆32/∆32) are resistant to infection by CCR5-using ("R5") HIV-1 strains but remain susceptible to less common CXCR4-using ("X4") strains. The evolutionary dynamics of X4 infections however, remain incompletely understood. We identified two individuals, one CCR5wt/wt and one CCR5∆32/∆32, within the Vancouver Injection Drug Users Study who were infected with a genetically similar X4 HIV-1 strain. While early-stage plasma viral loads were comparable in the two individuals (~4.5-5 log10 HIV-1 RNA copies/ml), CD4 counts in the CCR5wt/wt individual reached a nadir of <20 CD4 cells/mm(3) within 17 months but remained >250 cells/mm(3) in the CCR5∆32/∆32 individual. Ancestral phylogenetic reconstructions using longitudinal envelope-V3 deep sequences suggested that both individuals were infected by a single transmitted/founder (T/F) X4 virus that differed at only one V3 site (codon 24). While substantial within-host HIV-1 V3 diversification was observed in plasma and PBMC in both individuals, the CCR5wt/wt individual's HIV-1 population gradually reverted from 100% X4 to ~60% R5 over ~4 years whereas the CCR5∆32/∆32 individual's remained consistently X4. Our observations illuminate early dynamics of X4 HIV-1 infections and underscore the influence of CCR5 genotype on HIV-1 V3 evolution.

Decreasing seagrass density negatively influences associated fauna.

Seagrass meadows globally are disappearing at a rapid rate with physical disturbances being one of the major drivers of this habitat loss. Disturbance of seagrass can lead to fragmentation, a reduction in shoot density, canopy height and coverage, and potentially permanent loss of habitat. Despite being such a widespread issue, knowledge of how such small scale change affects the spatial distribution and abundances of motile fauna remains limited. The present study investigated fish and macro faunal community response patterns to a range of habitat variables (shoot length, cover and density), including individual species habitat preferences within a disturbed and patchy intertidal seagrass meadow. Multivariate analysis showed a measurable effect of variable seagrass cover on the abundance and distribution of the fauna, with species specific preferences to both high and low seagrass cover seagrass. The faunal community composition varied significantly with increasing/decreasing cover. The faunal species composition of low cover seagrass was more similar to sandy control plots than to higher cover seagrass. Shannon Wiener Diversity (H') and species richness was significantly higher in high cover seagrass than in low cover seagrass, indicating increasing habitat value as density increases. The results of this study underline how the impacts of small scale disturbances from factors such as anchor damage, boat moorings and intertidal vehicle use on seagrass meadows that reduce shoot density and cover can impact upon associated fauna. These impacts have negative consequences for the delivery of ecosystem services such as the provision of nursery habitat.

Global origin and transmission of hepatitis C virus nonstructural protein 3 Q80K polymorphism.

Hepatitis C virus (HCV) has a naturally occurring polymorphism, Q80K, in the nonstructural protein 3 (NS3) gene encoding the viral protease, which has been associated with reduced susceptibility to the direct-acting antiviral inhibitor simeprevir. Q80K is observed predominantly in HCV genotype 1a and seldom in other HCV genotypes; moreover, it has a markedly high prevalence in the United States. Here, we reconstruct the evolutionary history of this polymorphism to investigate why it is so highly localized in prevalence and whether it is stably transmitted between hosts. We found that the majority (96%) of HCV infections carrying Q80K were descended from a single lineage in which a Q80K substitution occurred around the 1940s in the United States, which implies that this polymorphism is likely highly transmissible. Furthermore, we identified 2 other substitutions in NS3 that may interact with Q80K and contribute to its stability. Our results imply that the current distribution and prevalence of Q80K are unlikely to change significantly in the short term.

Theoretical and experimental assessment of degenerate primer tagging in ultra-deep applications of next-generation sequencing.

Primer IDs (pIDs) are random oligonucleotide tags used in next-generation sequencing to identify sequences that originate from the same template. These tags are produced by degenerate primers during the reverse transcription of RNA molecules into cDNA. The use of pIDs helps to track the number of RNA molecules carried through amplification and sequencing, and allows resolution of inconsistencies between reads sharing a pID. Three potential issues complicate the above applications. First, multiple cDNAs may share a pID by chance; we found that while preventing any cDNAs from sharing a pID may be unfeasible, it is still practical to limit the number of these collisions. Secondly, a pID must be observed in at least three sequences to allow error correction; as such, pIDs observed only one or two times must be rejected. If the sequencing product contains copies from a high number of RT templates but produces few reads, our findings indicate that rejecting such pIDs will discard a great deal of data. Thirdly, the use of pIDs could influence amplification and sequencing. We examined the effects of several intrinsic and extrinsic factors on sequencing reads at both the individual and ensemble level.

An evaluation of phylogenetic methods for reconstructing transmitted HIV variants using longitudinal clonal HIV sequence data.

UNLABELLED: A population of human immunodeficiency virus (HIV) within a host often descends from a single transmitted/founder virus. The high mutation rate of HIV, coupled with long delays between infection and diagnosis, make isolating and characterizing this strain a challenge. In theory, ancestral reconstruction could be used to recover this strain from sequences sampled in chronic infection; however, the accuracy of phylogenetic techniques in this context is unknown. To evaluate the accuracy of these methods, we applied ancestral reconstruction to a large panel of published longitudinal clonal and/or single-genome-amplification HIV sequence data sets with at least one intrapatient sequence set sampled within 6 months of infection or seroconversion (n = 19,486 sequences, median [interquartile range] = 49 [20 to 86] sequences/set). The consensus of the earliest sequences was used as the best possible estimate of the transmitted/founder. These sequences were compared to ancestral reconstructions from sequences sampled at later time points using both phylogenetic and phylogeny-naive methods. Overall, phylogenetic methods conferred a 16% improvement in reproducing the consensus of early sequences, compared to phylogeny-naive methods. This relative advantage increased with intrapatient sequence diversity (P < 10(-5)) and the time elapsed between the earliest and subsequent samples (P < 10(-5)). However, neither approach performed well for reconstructing ancestral indel variation, especially within indel-rich regions of the HIV genome. Although further improvements are needed, our results indicate that phylogenetic methods for ancestral reconstruction significantly outperform phylogeny-naive alternatives, and we identify experimental conditions and study designs that can enhance accuracy of transmitted/founder virus reconstruction. IMPORTANCE: When HIV is transmitted into a new host, most of the viruses fail to infect host cells. Consequently, an HIV infection tends to be descended from a single "founder" virus. A priority target for the vaccine research, these transmitted/founder viruses are difficult to isolate since newly infected individuals are often unaware of their status for months or years, by which time the virus population has evolved substantially. Here, we report on the potential use of evolutionary methods to reconstruct the genetic sequence of the transmitted/founder virus from its descendants at later stages of an infection. These methods can recover this ancestral sequence with an overall error rate of about 2.3%-about 15% more information than if we had ignored the evolutionary relationships among viruses. Although there is no substitute for sampling infections at earlier points in time, these methods can provide useful information about the genetic makeup of transmitted/founder HIV.

Mapping the shapes of phylogenetic trees from human and zoonotic RNA viruses.

A phylogeny is a tree-based model of common ancestry that is an indispensable tool for studying biological variation. Phylogenies play a special role in the study of rapidly evolving populations such as viruses, where the proliferation of lineages is constantly being shaped by the mode of virus transmission, by adaptation to immune systems, and by patterns of human migration and contact. These processes may leave an imprint on the shapes of virus phylogenies that can be extracted for comparative study; however, tree shapes are intrinsically difficult to quantify. Here we present a comprehensive study of phylogenies reconstructed from 38 different RNA viruses from 12 taxonomic families that are associated with human pathologies. To accomplish this, we have developed a new procedure for studying phylogenetic tree shapes based on the 'kernel trick', a technique that maps complex objects into a statistically convenient space. We show that our kernel method outperforms nine different tree balance statistics at correctly classifying phylogenies that were simulated under different evolutionary scenarios. Using the kernel method, we observe patterns in the distribution of RNA virus phylogenies in this space that reflect modes of transmission and pathogenesis. For example, viruses that can establish persistent chronic infections (such as HIV and hepatitis C virus) form a distinct cluster. Although the visibly 'star-like' shape characteristic of trees from these viruses has been well-documented, we show that established methods for quantifying tree shape fail to distinguish these trees from those of other viruses. The kernel approach presented here potentially represents an important new tool for characterizing the evolution and epidemiology of RNA viruses.

Ability of HIV-1 Nef to downregulate CD4 and HLA class I differs among viral subtypes.

BACKGROUND: The highly genetically diverse HIV-1 group M subtypes may differ in their biological properties. Nef is an important mediator of viral pathogenicity; however, to date, a comprehensive inter-subtype comparison of Nef in vitro function has not been undertaken. Here, we investigate two of Nef's most well-characterized activities, CD4 and HLA class I downregulation, for clones obtained from 360 chronic patients infected with HIV-1 subtypes A, B, C or D. RESULTS: Single HIV-1 plasma RNA Nef clones were obtained from N=360 antiretroviral-naïve, chronically infected patients from Africa and North America: 96 (subtype A), 93 (B), 85 (C), and 86 (D). Nef clones were expressed by transfection in an immortalized CD4+ T-cell line. CD4 and HLA class I surface levels were assessed by flow cytometry. Nef expression was verified by Western blot. Subset analyses and multivariable linear regression were used to adjust for differences in age, sex and clinical parameters between cohorts. Consensus HIV-1 subtype B and C Nef sequences were synthesized and functionally assessed. Exploratory sequence analyses were performed to identify potential genotypic correlates of Nef function. Subtype B Nef clones displayed marginally greater CD4 downregulation activity (p = 0.03) and markedly greater HLA class I downregulation activity (p < 0.0001) than clones from other subtypes. Subtype C Nefs displayed the lowest in vitro functionality. Inter-subtype differences in HLA class I downregulation remained statistically significant after controlling for differences in age, sex, and clinical parameters (p < 0.0001). The synthesized consensus subtype B Nef showed higher activities compared to consensus C Nef, which was most pronounced in cells expressing lower protein levels. Nef clones exhibited substantial inter-subtype diversity: cohort consensus residues differed at 25% of codons, while a similar proportion of codons exhibited substantial inter-subtype differences in major variant frequency. These amino acids, along with others identified in intra-subtype analyses, represent candidates for mediating inter-subtype differences in Nef function. CONCLUSIONS: Results support a functional hierarchy of subtype B > A/D > C for Nef-mediated CD4 and HLA class I downregulation. The mechanisms underlying these differences and their relevance to HIV-1 pathogenicity merit further investigation.