Lewi Stone

Multiannual forecasting of seasonal influenza dynamics reveals climatic and evolutionary drivers

Significance Seasonality in disease incidence is ubiquitous among infectious diseases. Seasonal drivers include weather variables, such as temperature and humidity, and social factors (e.g., contact patterns). Attempts to make long-term predictions of infectious diseases are hampered by the inability to understand the complex interplay of these factors. Here, we model the dynamics of seasonal influenza based on a high-quality 12-year Israeli dataset. The dynamics are completely explainable by the time evolution of the model equations, the antigenic jumps of the virus, and the climatic forcing, yielding high-correlation fits to the data (r = 0.94). The forecasting ability is greatly increased through the predictability of the system’s transient dynamics, resulting in accurate predictions (r = 0.93) that have not yet been found elsewhere. Human influenza occurs annually in most temperate climatic zones of the world, with epidemics peaking in the cold winter months. Considerable debate surrounds the relative role of epidemic dynamics, viral evolution, and climatic drivers in driving year-to-year variability of outbreaks. The ultimate test of understanding is prediction; however, existing influenza models rarely forecast beyond a single year at best. Here, we use a simple epidemiological model to reveal multiannual predictability based on high-quality influenza surveillance data for Israel; the model fit is corroborated by simple metapopulation comparisons within Israel. Successful forecasts are driven by temperature, humidity, antigenic drift, and immunity loss. Essentially, influenza dynamics are a balance between large perturbations following significant antigenic jumps, interspersed with nonlinear epidemic dynamics tuned by climatic forcing.

Modelling seasonal influenza: the role of weather and punctuated antigenic drift

Seasonal influenza appears as annual oscillations in temperate regions of the world, yet little is known as to what drives these annual outbreaks and what factors are responsible for their inter-annual variability. Recent studies suggest that weather variables, such as absolute humidity, are the key drivers of annual influenza outbreaks. The rapid, punctuated, antigenic evolution of the influenza virus is another major factor. We present a new framework for modelling seasonal influenza based on a discrete-time, age-of-infection, epidemic model, which allows the calculation of the models likelihood function in closed form. This framework may be used to perform model inference and parameter estimation rigorously. The modelling approach allows us to fit 11 years of Israeli influenza data, with the best models fitting the data with unusually high correlations in which r > 0.9. We show that using actual weather to modulate influenza transmission rate gives better results than using the inter-annual means of the weather variables, providing strong support for the role of weather in shaping the dynamics of influenza. This conclusion remains valid even when incorporating a more realistic depiction of the decay of immunity at the population level, which allows for discrete changes in immunity from year to year.

Global Spatio-temporal Patterns of Influenza in the Post-pandemic Era

We study the global spatio-temporal patterns of influenza dynamics. This is achieved by analysing and modelling weekly laboratory confirmed cases of influenza A and B from 138 countries between January 2006 and January 2015. The data were obtained from FluNet, the surveillance network compiled by the the World Health Organization. We report a pattern of skip-and-resurgence behavior between the years 2011 and 2013 for influenza H1N1pdm, the strain responsible for the 2009 pandemic, in Europe and Eastern Asia. In particular, the expected H1N1pdm epidemic outbreak in 2011/12 failed to occur (or “skipped”) in many countries across the globe, although an outbreak occurred in the following year. We also report a pattern of well-synchronized wave of H1N1pdm in early 2011 in the Northern Hemisphere countries, and a pattern of replacement of strain H1N1pre by H1N1pdm between the 2009 and 2012 influenza seasons. Using both a statistical and a mechanistic mathematical model, and through fitting the data of 108 countries, we discuss the mechanisms that are likely to generate these events taking into account the role of multi-strain dynamics. A basic understanding of these patterns has important public health implications and scientific significance.

Modeling the Impact of White-Plague Coral Disease in Climate Change Scenarios.

Coral reefs are in global decline, with coral diseases increasing both in prevalence and in space, a situation that is expected only to worsen as future thermal stressors increase. Through intense surveillance, we have collected a unique and highly resolved dataset from the coral reef of Eilat (Israel, Red Sea), that documents the spatiotemporal dynamics of a White Plague Disease (WPD) outbreak over the course of a full season. Based on modern statistical methodologies, we develop a novel spatial epidemiological model that uses a maximum-likelihood procedure to fit the data and assess the transmission pattern of WPD. We link the model to sea surface temperature (SST) and test the possible effect of increasing temperatures on disease dynamics. Our results reveal that the likelihood of a susceptible coral to become infected is governed both by SST and by its spatial location relative to nearby infected corals. The model shows that the magnitude of WPD epidemics strongly depends on demographic circumstances; under one extreme, when recruitment is free-space regulated and coral density remains relatively constant, even an increase of only 0.5°C in SST can cause epidemics to double in magnitude. In reality, however, the spatial nature of transmission can effectively protect the community, restricting the magnitude of annual epidemics. This is because the probability of susceptible corals to become infected is negatively associated with coral density. Based on our findings, we expect that infectious diseases having a significant spatial component, such as Red-Sea WPD, will never lead to a complete destruction of the coral community under increased thermal stress. However, this also implies that signs of recovery of local coral communities may be misleading; indicative more of spatial dynamics than true rehabilitation of these communities. In contrast to earlier generic models, our approach captures dynamics of WPD both in space and time, accounting for the highly seasonal nature of annual WPD outbreaks.

Finding the Most Influential Nodes in Pinning Controllability of Complex Networks

Identifying the best drivers (i.e., the nodes to apply the control signals in a large complex network), which gives the fastest synchronization to the reference state, is a challenge in pinning control of a network. There is not yet a method that exactly predicts a set of best drivers. In this brief, we introduce a novel method that gives first-order approximation for the importance of nodes in pinning control. A spectral measure (the largest eigenvalue of the augmented Laplacian matrix divided by the smallest eigenvalue) is considered as a pinning controllability metric. We develop this method based on the sensitivity analysis of the Laplacian eigenratio, resulting in the scoring of nodes based on their importance in pinning control. The method is rather simple to compute and needs a single eigendecomposition of the Laplacian matrix of the connection graph. Applying this technique on a number of model networks reveals its effectiveness over heuristic approaches.

Integrated species distribution models: combining presence‐background data and site‐occupancy data with imperfect detection

Summary Two main sources of data for species distribution models (SDMs) are site-occupancy (SO) data from planned surveys, and presence-background (PB) data from opportunistic surveys and other sources. SO surveys give high quality data about presences and absences of the species in a particular area. However, due to their high cost, they often cover a smaller area relative to PB data, and are usually not representative of the geographic range of a species. In contrast, PB data is plentiful, covers a larger area, but is less reliable due to the lack of information on species absences, and is usually characterised by biased sampling. Here we present a new approach for species distribution modelling that integrates these two data types. We have used an inhomogeneous Poisson point process as the basis for constructing an integrated SDM that fits both PB and SO data simultaneously. It is the first implementation of an Integrated SO–PB Model which uses repeated survey occupancy data and also incorporates detection probability. The Integrated Models performance was evaluated, using simulated data and compared to approaches using PB or SO data alone. It was found to be superior, improving the predictions of species spatial distributions, even when SO data is sparse and collected in a limited area. The Integrated Model was also found effective when environmental covariates were significantly correlated. Our method was demonstrated with real SO and PB data for the Yellow-bellied glider (Petaurus australis) in south-eastern Australia, with the predictive performance of the Integrated Model again found to be superior. PB models are known to produce biased estimates of species occupancy or abundance. The small sample size of SO datasets often results in poor out-of-sample predictions. Integrated models combine data from these two sources, providing superior predictions of species abundance compared to using either data source alone. Unlike conventional SDMs which have restrictive scale-dependence in their predictions, our Integrated Model is based on a point process model and has no such scale-dependency. It may be used for predictions of abundance at any spatial-scale while still maintaining the underlying relationship between abundance and area.

The Google matrix controls the stability of structured ecological and biological networks

Mays celebrated theoretical work of the 70s contradicted the established paradigm by demonstrating that complexity leads to instability in biological systems. Here Mays random-matrix modelling approach is generalized to realistic large-scale webs of species interactions, be they structured by networks of competition, mutualism or both. Simple relationships are found to govern these otherwise intractable models, and control the parameter ranges for which biological systems are stable and feasible. Our analysis of model and real empirical networks is only achievable on introducing a simplifying Google-matrix reduction scheme, which in the process, yields a practical ecological eigenvalue stability index. These results provide an insight into how network topology, especially connectance, influences species stable coexistence. Constraints controlling feasibility (positive equilibrium populations) in these systems are found more restrictive than those controlling stability, helping explain the enigma of why many classes of feasible ecological models are nearly always stable.

The effects of connectivity on metapopulation persistence: network symmetry and degree correlations

A spatial metapopulation is a mosaic of interconnected patch populations. The complex routes of colonization between the patches are governed by the metapopulations dispersal network. Over the past two decades, there has been considerable interest in uncovering the effects of dispersal network topology and its symmetry on metapopulation persistence. While most studies find that the level of symmetry in dispersal pattern enhances persistence, some have reached the conclusion that symmetry has at most a minor effect. In this work, we present a new perspective on the debate. We study properties of the in- and out-degree distribution of patches in the metapopulation which define the number of dispersal routes into and out of a particular patch, respectively. By analysing the spectral radius of the dispersal matrices, we confirm that a higher level of symmetry has only a marginal impact on persistence. We continue to analyse different properties of the in–out degree distribution, namely the ‘in–out degree correlation’ (IODC) and degree heterogeneity, and find their relationship to metapopulation persistence. Our analysis shows that, in contrast to symmetry, the in–out degree distribution and particularly, the IODC are dominant factors controlling persistence.

Analysis of Influenza and RSV dynamics in the community using a ‘Local Transmission Zone’ approach

Understanding the dynamics of pathogen spread within urban areas is critical for the effective prevention and containment of communicable diseases. At these relatively small geographic scales, short-distance interactions and tightly knit sub-networks dominate the dynamics of pathogen transmission; yet, the effective boundaries of these micro-scale groups are generally not known and often ignored. Using clinical test results from hospital admitted patients we analyze the spatio-temporal distribution of Influenza Like Illness (ILI) in the city of Jerusalem over a period of three winter seasons. We demonstrate that this urban area is not a single, perfectly mixed ecology, but is in fact comprised of a set of more basic, relatively independent pathogen transmission units, which we term here Local Transmission Zones, LTZs. By identifying these LTZs, and using the dynamic pathogen-content information contained within them, we are able to differentiate between disease-causes at the individual patient level often with near-perfect predictive accuracy.

A novel method for estimating the number of species within a region

Ecologists are often required to estimate the number of species in a region or designated area. A number of diversity indices are available for this purpose and are based on sampling the area using quadrats or other means, and estimating the total number of species from these samples. In this paper, a novel theory and method for estimating the number of species is developed. The theory involves the use of the Laplace method for approximating asymptotic integrals. The method is shown to be successful by testing random simulated datasets. In addition, several real survey datasets are tested, including forests that contain a large number (tens to hundreds) of tree species, and an aquatic system with a large number of fish species. The method is shown to give accurate results, and in almost all cases found to be superior to existing tools for estimating diversity.