A coupled hidden Markov model for disease interactions

Patterns and drivers of vector-borne microparasites in a classic metapopulation

Many organisms live in fragmented populations, which has profound consequences on the dynamics of associated parasites. Metapopulation theory offers a canonical framework for predicting the effects of fragmentation on spatiotemporal host–parasite dynamics. However, empirical studies of parasites in classical metapopulations remain rare, particularly for vector-borne parasites. Here, we quantify spatiotemporal patterns and possible drivers of infection probability for several ectoparasites (fleas, Ixodes trianguliceps and Ixodes ricinus) and vector-borne microparasites (Babesia microti, Bartonella spp., Hepatozoon spp.) in a classically functioning metapopulation of water vole hosts. Results suggest that the relative importance of vector or host dynamics on microparasite infection probabilities is related to parasite life-histories. Bartonella, a microparasite with a fast life-history, was positively associated with both host and vector abundances at several spatial and temporal scales. In contrast, B. microti, a tick-borne parasite with a slow life-history, was only associated with vector dynamics. Further, we provide evidence that life-history shaped parasite dynamics, including occupancy and colonization rates, in the metapopulation. Lastly, our findings were consistent with the hypothesis that landscape connectivity was determined by distance-based dispersal of the focal hosts. We provide essential empirical evidence that contributes to the development of a comprehensive theory of metapopulation processes of vector-borne parasites.

Modelling methicillin-resistant Staphylococcus aureus decolonization: interactions between body sites and the impact of site-specific clearance

Methicillin-resistant Staphylococcus aureus (MRSA) can colonize multiple body sites, and carriage is a risk factor for infection. Successful decolonization protocols reduce disease incidence; however, multiple protocols exist, comprising diverse therapies targeting multiple body sites, and the optimal protocol is unclear. Standard methods cannot infer the impact of site-specific components on successful decolonization. Here, we formulate a Bayesian coupled hidden Markov model, which estimates interactions between body sites, quantifies the contribution of each therapy to successful decolonization, and enables predictions of the efficacy of therapy combinations. We applied the model to longitudinal data from a randomized controlled trial (RCT) of an MRSA decolonization protocol consisting of chlorhexidine body and mouthwash and nasal mupirocin. Our findings (i) confirmed nares as a central hub for MRSA colonization and nasal mupirocin as the most crucial therapy and (ii) demonstrated all components contributed significantly to the efficacy of the protocol and the protocol reduced self-inoculation. Finally, we assessed the impact of hypothetical therapy improvements in silico and found that enhancing MRSA clearance at the skin would yield the largest gains. This study demonstrates the use of advanced modelling to go beyond what is typically achieved by RCTs, enabling evidence-based decision-making to streamline clinical protocols.

A copula-based multivariate hidden Markov model for modelling momentum in football

We investigate the potential occurrence of change points—commonly referred to as “momentum shifts”—in the dynamics of football matches. For that purpose, we model minute-by-minute in-game statistics of Bundesliga matches using hidden Markov models (HMMs). To allow for within-state dependence of the variables, we formulate multivariate state-dependent distributions using copulas. For the Bundesliga data considered, we find that the fitted HMMs comprise states which can be interpreted as a team showing different levels of control over a match. Our modelling framework enables inference related to causes of momentum shifts and team tactics, which is of much interest to managers, bookmakers, and sports fans.

Uncovering ecological state dynamics with hidden Markov models

Ecological systems can often be characterised by changes among a finite set of underlying states pertaining to individuals, populations, communities or entire ecosystems through time. Owing to the inherent difficulty of empirical field studies, ecological state dynamics operating at any level of this hierarchy can often be unobservable or 'hidden'. Ecologists must therefore often contend with incomplete or indirect observations that are somehow related to these underlying processes. By formally disentangling state and observation processes based on simple yet powerful mathematical properties that can be used to describe many ecological phenomena, hidden Markov models (HMMs) can facilitate inferences about complex system state dynamics that might otherwise be intractable. However, HMMs have only recently begun to gain traction within the broader ecological community. We provide a gentle introduction to HMMs, establish some common terminology, review the immense scope of HMMs for applied ecological research and provide a tutorial on implementation and interpretation. By illustrating how practitioners can use a simple conceptual template to customise HMMs for their specific systems of interest, revealing methodological links between existing applications, and highlighting some practical considerations and limitations of these approaches, our goal is to help establish HMMs as a fundamental inferential tool for ecologists.

A primer on coupled state-switching models for multiple interacting time series

State-switching models such as hidden Markov models or Markov-switching regression models are routinely applied to analyse sequences of observations that are driven by underlying non-observable states. Coupled state-switching models extend these approaches to address the case of multiple observation sequences whose underlying state variables interact. In this article, we provide an overview of the modelling techniques related to coupling in state-switching models, thereby forming a rich and flexible statistical framework particularly useful for modelling correlated time series. Simulation experiments demonstrate the relevance of being able to account for an asynchronous evolution as well as interactions between the underlying latent processes. The models are further illustrated using two case studies related to (a) interactions between a dolphin mother and her calf as inferred from movement data and (b) electronic health record data collected on 696 patients within an intensive care unit.

Scalable Bayesian Inference for Coupled Hidden Markov and Semi-Markov Models

Bayesian inference for coupled hidden Markov models frequently relies on data augmentation techniques for imputation of the hidden state processes. Considerable progress has been made on developing such techniques, mainly using Markov chain Monte Carlo (MCMC) methods. However, as the dimensionality and complexity of the hidden processes increase some of these methods become inefficient, either because they produce MCMC chains with high autocorrelation or because they become computationally intractable. Motivated by this fact we developed a novel MCMC algorithm, which is a modification of the forward filtering backward sampling algorithm, that achieves a good balance between computation and mixing properties, and thus can be used to analyze models with large numbers of hidden chains. Even though our approach is developed under the assumption of a Markovian model, we show how this assumption can be relaxed leading to minor modifications in the algorithm. Our approach is particularly well suited to epidemic models, where the hidden Markov chains represent the infection status of an individual through time. The performance of our method is assessed on simulated data on epidemic models for the spread of Escherichia coli O157:H7 in cattle. Supplementary materials for this article are available online.

Variational Inference for Coupled Hidden Markov Models Applied to the Joint Detection of Copy Number Variations

Hidden Markov models provide a natural statistical framework for the detection of the copy number variations (CNV) in genomics. In this context, we define a hidden Markov process that underlies all individuals jointly in order to detect and to classify genomics regions in different states (typically, deletion, normal or amplification). Structural variations from different individuals may be dependent. It is the case in agronomy where varietal selection program exists and species share a common phylogenetic past. We propose to take into account these dependencies inthe HMM model. When dealing with a large number of series, maximum likelihood inference (performed classically using the EM algorithm) becomes intractable. We thus propose an approximate inference algorithm based on a variational approach (VEM), implemented in the CHMM R package. A simulation study is performed to assess the performance of the proposed method and an application to the detection of structural variations in plant genomes is presented.

Physiological, but not fitness, effects of two interacting haemoparasitic infections in a wild rodent

In contrast to the conditions in most laboratory studies, wild animals are routinely challenged by multiple infections simultaneously, and these infections can interact in complex ways. This means that the impact of a parasite on its host's physiology and fitness cannot be fully assessed in isolation, and requires consideration of the interactions with other co-infections. Here we examine the impact of two common blood parasites in the field vole (Microtus agrestis): Babesia microti and Bartonella spp., both of which have zoonotic potential. We collected longitudinal and cross-sectional data from four populations of individually tagged wild field voles. This included data on biometrics, life history, ectoparasite counts, presence/absence of microparasites, immune markers and, for a subset of voles, more detailed physiological and immunological measurements. This allowed us to monitor infections over time and to estimate components of survival and fecundity. We confirm, as reported previously, that B. microti has a preventative effect on infection with Bartonella spp., but that the reverse is not true. We observed gross splenomegaly following B. microti infection, and an increase in IL-10 production together with some weight loss following Bartonella spp. infection. However, these animals appeared otherwise healthy and we detected no impact of infection on survival or fecundity due to the two haemoparasite taxa. This is particularly remarkable in the case of B. microti which induces apparently drastic long-term changes to spleen sizes, but without major adverse effects. Our work sheds light on the ecologies of these important zoonotic agents, and more generally on the influence that interactions among multiple parasites have on their hosts in the wild.

A Profile Hidden Markov Model to investigate the distribution and frequency of LanB-encoding lantibiotic modification genes in the human oral and gut microbiome

Background

The human microbiota plays a key role in health and disease, and bacteriocins, which are small, bacterially produced, antimicrobial peptides, are likely to have an important function in the stability and dynamics of this community. Here we examined the density and distribution of the subclass I lantibiotic modification protein, LanB, in human oral and stool microbiome datasets using a specially constructed profile Hidden Markov Model (HMM).

Methods

The model was validated by correctly identifying known lanB genes in the genomes of known bacteriocin producers more effectively than other methods, while being sensitive enough to differentiate between different subclasses of lantibiotic modification proteins. This approach was compared with two existing methods to screen both genomic and metagenomic datasets obtained from the Human Microbiome Project (HMP).

Results

Of the methods evaluated, the new profile HMM identified the greatest number of putative LanB proteins in the stool and oral metagenome data while BlastP identified the fewest. In addition, the model identified more LanB proteins than a pre-existing Pfam lanthionine dehydratase model. Searching the gastrointestinal tract subset of the HMP reference genome database with the new HMM identified seven putative subclass I lantibiotic producers, including two members of the Coprobacillus genus.

Conclusions

These findings establish custom profile HMMs as a potentially powerful tool in the search for novel bioactive producers with the power to benefit human health, and reinforce the repertoire of apparent bacteriocin-encoding gene clusters that may have been overlooked by culture-dependent mining efforts to date.

The importance of multiparasitism: examining the consequences of co-infections for human and animal health

Most parasites co-occur with other parasites, although the importance of such multiparasitism has only recently been recognised. Co-infections may result when hosts are independently infected by different parasites at the same time or when interactions among parasite species facilitate co-occurrence. Such interactions can have important repercussions on human or animal health because they can alter host susceptibility, infection duration, transmission risks, and clinical symptoms. These interactions may be synergistic or antagonistic and thus produce diverse effects in infected humans and animals. Interactions among parasites strongly influence parasite dynamics and therefore play a major role in structuring parasite populations (both within and among hosts) as well as host populations. However, several methodological challenges remain when it comes to detecting parasite interactions. The goal of this review is to summarise current knowledge on the causes and consequences of multiparasitism and to discuss the different methods and tools that researchers have developed to study the factors that lead to multiparasitism. It also identifies new research directions to pursue.

Host-parasite biology in the real world: the field voles of Kielder

Research on the interactions between the field voles (Microtus agrestis) of Kielder Forest and their natural parasites dates back to the 1930s. These early studies were primarily concerned with understanding how parasites shape the characteristic cyclic population dynamics of their hosts. However, since the early 2000s, research on the Kielder field voles has expanded considerably and the system has now been utilized for the study of host-parasite biology across many levels, including genetics, evolutionary ecology, immunology and epidemiology. The Kielder field voles therefore represent one of the most intensely and broadly studied natural host-parasite systems, bridging theoretical and empirical approaches to better understand the biology of infectious disease in the real world. This article synthesizes the body of work published on this system and summarizes some important insights and general messages provided by the integrated and multidisciplinary study of host-parasite interactions in the natural environment.

To be or not to be associated: power study of four statistical modeling approaches to identify parasite associations in cross-sectional studies

A growing number of studies are reporting simultaneous infections by parasites in many different hosts. The detection of whether these parasites are significantly associated is important in medicine and epidemiology. Numerous approaches to detect associations are available, but only a few provide statistical tests. Furthermore, they generally test for an overall detection of association and do not identify which parasite is associated with which other one. Here, we developed a new approach, the association screening approach, to detect the overall and the detail of multi-parasite associations. We studied the power of this new approach and of three other known ones (i.e., the generalized chi-square, the network and the multinomial GLM approaches) to identify parasite associations either due to parasite interactions or to confounding factors. We applied these four approaches to detect associations within two populations of multi-infected hosts: (1) rodents infected with Bartonella sp., Babesia microti and Anaplasma phagocytophilum and (2) bovine population infected with Theileria sp. and Babesia sp. We found that the best power is obtained with the screening model and the generalized chi-square test. The differentiation between associations, which are due to confounding factors and parasite interactions was not possible. The screening approach significantly identified associations between Bartonella doshiae and B. microti, and between T. parva, T. mutans, and T. velifera. Thus, the screening approach was relevant to test the overall presence of parasite associations and identify the parasite combinations that are significantly over- or under-represented. Unraveling whether the associations are due to real biological interactions or confounding factors should be further investigated. Nevertheless, in the age of genomics and the advent of new technologies, it is a considerable asset to speed up researches focusing on the mechanisms driving interactions between parasites.