Bayesian phylodynamics of avian influenza A virus H9N2 in Asia with time-dependent predictors of migration.

Model-based phylodynamic approaches recently employed generalized linear models (GLMs) to uncover potential predictors of viral spread. Very recently some of these models have allowed both the predictors and their coefficients to be time-dependent. However, these studies mainly focused on predictors that are assumed to be constant through time. Here we inferred the phylodynamics of avian influenza A virus H9N2 isolated in 12 Asian countries and regions under both discrete trait analysis (DTA) and structured coalescent (MASCOT) approaches. Using MASCOT we applied a new time-dependent GLM to uncover the underlying factors behind H9N2 spread. We curated a rich set of time-series predictors including annual international live poultry trade and national poultry production figures. This time-dependent phylodynamic prediction model was compared to commonly employed time-independent alternatives. Additionally the time-dependent MASCOT model allowed for the estimation of viral effective sub-population sizes and their changes through time, and these effective population dynamics within each country were predicted by a GLM. International annual poultry trade is a strongly supported predictor of virus migration rates. There was also strong support for geographic proximity as a predictor of migration rate in all GLMs investigated. In time-dependent MASCOT models, national poultry production was also identified as a predictor of virus genetic diversity through time and this signal was obvious in mainland China. Our application of a recently introduced time-dependent GLM predictors integrated rich time-series data in Bayesian phylodynamic prediction. We demonstrated the contribution of poultry trade and geographic proximity (potentially unheralded wild bird movements) to avian influenza spread in Asia. To gain a better understanding of the drivers of H9N2 spread, we suggest increased surveillance of the H9N2 virus in countries that are currently under-sampled as well as in wild bird populations in the most affected countries.

The Structured Coalescent and Its Approximations.

Phylogeographic methods can help reveal the movement of genes between populations of organisms. This has been widely done to quantify pathogen movement between different host populations, the migration history of humans, and the geographic spread of languages or gene flow between species using the location or state of samples alongside sequence data. Phylogenies therefore offer insights into migration processes not available from classic epidemiological or occurrence data alone. Phylogeographic methods have however several known shortcomings. In particular, one of the most widely used methods treats migration the same as mutation, and therefore does not incorporate information about population demography. This may lead to severe biases in estimated migration rates for data sets where sampling is biased across populations. The structured coalescent on the other hand allows us to coherently model the migration and coalescent process, but current implementations struggle with complex data sets due to the need to infer ancestral migration histories. Thus, approximations to the structured coalescent, which integrate over all ancestral migration histories, have been developed. However, the validity and robustness of these approximations remain unclear. We present an exact numerical solution to the structured coalescent that does not require the inference of migration histories. Although this solution is computationally unfeasible for large data sets, it clarifies the assumptions of previously developed approximate methods and allows us to provide an improved approximation to the structured coalescent. We have implemented these methods in BEAST2, and we show how these methods compare under different scenarios.

Reconstructing the history of residence strategies in Indo-European-speaking societies: neo-, uxori-, and virilocality.

Linguists and archaeologists have used reconstructions of early Indo-European residence strategies to constrain hypotheses about the homeland and trajectory of dispersal of Indo-European languages; however, these reconstructions are largely based on unsystematic and a historical use of the linguistic and ethnographic evidence, coupled with substantial bias in interpretation. Here I use cross-cultural data in a phylogenetic comparative framework to reconstruct the pattern of change in residence strategies in the history of societies speaking Indo-European languages. The analysis provides evidence in support of prevailing virilocality with alternative neolocality for Proto-Indo-European, and that this pattern may have extended back to Proto-Indo-Hittite. These findings bolster interpretations of the archaeological evidence that emphasize the "non-matricentric" structure of early Indo-European society; however, they also counter the notion that early Indo-European society was strongly "patricentric." I discuss implications of these findings in the context of the archaeological and genetic evidence on prehistoric social organization.

Genomic evidence for an African expansion of anatomically modern humans by a Southern route.

There is general agreement among scientists about a recent (less than 200,000 yrs ago) African origin of anatomically modern humans, whereas there is still uncertainty about whether, and to what extent, they admixed with archaic populations, which thus may have contributed to the modern populations' gene pools. Data on cranial morphology have been interpreted as suggesting that, before the main expansion from Africa through the Near East, anatomically modern humans may also have taken a Southern route from the Horn of Africa through the Arabian peninsula to India, Melanesia and Australia, about 100,000 yrs ago. This view was recently supported by archaeological findings demonstrating human presence in Eastern Arabia >90,000 yrs ago. In this study we analyzed genetic variation at 111,197 nuclear SNPs in nine populations (Kurumba, Chenchu, Kamsali, Madiga, Mala, Irula, Dalit, Chinese, Japanese), chosen because their genealogical relationships are expected to differ under the alternative models of expansion (single vs. multiple dispersals). We calculated correlations between genomic distances, and geographic distances estimated under the alternative assumptions of a single dispersal, or multiple dispersals, and found a significantly stronger association for the multiple dispersal model. If confirmed, this result would cast doubts on the possibility that some non-African populations (i.e., those whose ancestors expanded through the Southern route) may have had any contacts with Neandertals.

Geographic patterns of craniofacial variation in pre-Hispanic populations from the Southern Cone of South America.

In this study we analyzed the relationships and patterns of spatial variation from morphological cranial variability of 17 population samples representing the ancient inhabitants of the central territory of Argentina (archaeologically known as "Sierras Centrales") and other pre-Hispanic populations from different ecological and geographic regions of the Southern Cone of South America (Argentina and Uruguay), based on the analysis of 10 craniofacial measurements. Results obtained from D2 distances can be interpreted as evidence of a similar biological history for the populations that inhabited the Sierras Centrales and the population of Santiago del Estero. Matrix correlation analysis demonstrated that craniometric variation is significantly influenced by geography, suggesting that populations that lived at lower geographical distance share more biological similarity. Global spatial autocorrelation analysis suggests a clinal pattern for the biological variation, although Moran's I estimates calculated for each variable demonstrate that only nasal height and breadth show this spatial pattern of variation. Results from spatial regression techniques show a significant effect of altitude modeling nasal shape, in agreement with previous studies suggesting that nasal morphology is strongly influenced by environment variables.

Paternal genetic history of the Basque population of Spain.

This study examines the genetic variation in Basque Y chromosome lineages using data on 12 Y-short tandem repeat (STR) loci in a sample of 158 males from four Basque provinces of Spain (Alava, Vizcaya, Guipuzcoa, and Navarre). As reported in previous studies, the Basques are characterized by high frequencies of haplogroup R1b (83%). AMOVA analysis demonstrates genetic homogeneity, with a small but significant amount of genetic structure between provinces (Y-short tandem repeat loci STRs: 1.71%, p = 0.0369). Gene and haplotype diversity levels in the Basque population are on the low end of the European distribution (gene diversity: 0.4268; haplotype diversity: 0.9421). Post-Neolithic contribution to the paternal Basque gene pool was estimated by measuring the proportion of those haplogroups with a Time to Most Recent Common Ancestor (TMRCA) previously dated either prior (R1b, I2a2) or subsequent to (E1b1b, G2a, J2a) the Neolithic. Based on these estimates, the Basque provinces show varying degrees of post-Neolithic contribution in the paternal lineages (10.9% in the combined sample).

[Quantitative estimation of area parameters (with representatives of genus rana as a case study)].

A quantitative method of species "point" area analysis is considered that provides the interpolation of species distribution to the whole territory on the basis of its relationships with climatic and relief variables. It is shown that application of standart statistical interpolation techniques is incorrect. The proposed approach is based on interpolation onto the whole territory the species-specific relations with environmental variables detected in single "points". The basic method for the task proves to be the factor analysis. Within the scope of the study, we have considered the methods for quantitative representation of species relationships with climatic and relief variables. The analysis efficiency is demonstrated by an example of three species of brown frogs: Rana temporaria, R. arvalis u R. amurensis.

Spatial Congruence Analysis (SCAN): A method for detecting biogeographical patterns based on species range congruences.

Species with congruent geographical distributions, potentially caused by common historical and ecological spatial processes, constitute biogeographical units called chorotypes. Nevertheless, the degree of spatial range congruence characterizing these groups of species is rarely used as an explicit parameter. Methods conceived for the identification of patterns of shared ranges often suffer from scale bias associated with the use of grids, or the incapacity to describe the full complexity of patterns, from core areas of high spatial congruence, to long gradients of range distributions expanding from these core areas. Here, we propose a simple analytical method, Spatial Congruence Analysis (SCAN), which identifies chorotypes by mapping direct and indirect spatial relationships among species. Assessments are made under a referential value of congruence as an explicit numerical parameter. A one-layered network connects species (vertices) using pairwise spatial congruence estimates (edges). This network is then analyzed for each species, separately, by an algorithm which searches for spatial relationships to the reference species. The method was applied to two datasets: a simulated gradient of ranges and real distributions of birds. The simulated dataset showed that SCAN can describe gradients of distribution with a high level of detail. The bird dataset showed that only a small portion of range overlaps is biogeographically meaningful, and that there is a large variation in types of patterns that can be found with real distributions. Species analyzed separately may converge on similar or identical groups, may be nested in larger chorotypes, or may even generate overlapped patterns with no species in common. Chorotypes can vary from simple ones, composed by few highly congruent species, to complex, with numerous alternative component species and spatial configurations, which offer insights about possible processes driving these patterns in distinct degrees of spatial congruence. Metrics such as congruence, depth, richness, and ratio between common and total areas can be used to describe chorotypes in detail, allowing comparisons between patterns across regions and taxa.

Reconstructing unseen transmission events to infer dengue dynamics from viral sequences.

For most pathogens, transmission is driven by interactions between the behaviours of infectious individuals, the behaviours of the wider population, the local environment, and immunity. Phylogeographic approaches are currently unable to disentangle the relative effects of these competing factors. We develop a spatiotemporally structured phylogenetic framework that addresses these limitations by considering individual transmission events, reconstructed across spatial scales. We apply it to geocoded dengue virus sequences from Thailand (N = 726 over 18 years). We find infected individuals spend 96% of their time in their home community compared to 76% for the susceptible population (mainly children) and 42% for adults. Dynamic pockets of local immunity make transmission more likely in places with high heterotypic immunity and less likely where high homotypic immunity exists. Age-dependent mixing of individuals and vector distributions are not important in determining spread. This approach provides previously unknown insights into one of the most complex disease systems known and will be applicable to other pathogens.

South from Alaska: a pilot aDNA study of genetic history on the Alaska Peninsula and the eastern Aleutians.

The Aleutian Islands were colonized, perhaps several times, from the Alaskan mainland. Earlier work documented transitions in the relative frequencies of mtDNA haplogroups over time, but little is known about potential source populations for prehistoric Aleut migrants. As part of a pilot investigation, we sequenced the mtDNA first hypervariable region (HVRI) in samples from two archaeological sites on the Alaska Peninsula (the Hot Springs site near Port Moller, Alaska; and samples from a cluster of sites in the Brooks River area near Katmai National Park and Preserve) and one site from Prince William Sound (Mink Island). The sequences revealed not only the mtDNA haplogroups typically found in both ancient and modern Aleut populations (A2 and D2) but also haplogroups B2 and D1 in the Brooks River samples and haplogroup D3 in one Mink Islander. These preliminary results suggest greater mtDNA diversity in prehistoric populations than previously observed and facilitate reconstruction of migration scenarios from the peninsula into the Aleutian archipelago in the past.

Origins of Aleuts and the genetic structure of populations of the archipelago: molecular and archaeological perspectives.

We summarize the results of a field and laboratory research program (1999-2006) in the Aleutian Islands on the origins of the inhabitants of the archipelago and the genetic structure of these populations. The Aleuts show closest genetic affinity to the contemporary Siberian Eskimos and Chukchi of Chukotka and differ significantly from the populations of Kamchatka (the terminus of the archipelago) and Alaskan Eskimos. Our findings support the hypothesis that the ancestors of the Aleuts crossed Beringia and expanded westerly into the islands approximately 9,000 years ago. The Monmonier algorithm indicates genetic discontinuity between contemporary Kamchatkan populations and western Aleut populations, suggesting that island hopping from Kamchatka into the western Aleutian Islands was highly unlikely. The primary determinant of the distribution of genes throughout the archipelago is geography. The most intimate relationship exists between the genetics (based on mtDNA sequences and intermatch/mismatch distances) and geographic distances (measured in kilometers). However, the Y-chromosome haplogroup frequencies are not significantly correlated with the geography of the Aleutian Islands. The underlying patterns of precontact genetic structure based on Y-chromosome markers of the Aleut populations is obscured because of the gene flow from Russian male colonizers and Scandinavian and English fishermen. We consider alternative theories about the peopling of the Americas from Siberia. In addition, we attempt a synthesis between archaeological and genetic data for the Aleutian Islands.

Genetic architecture of a small, recently aggregated Aleut population: Bering Island, Russia.

The fishing community of Bering Island, located in the Russian Commander Islands off the Kamchatka Peninsula, was originally founded by a small number of Russian soldiers and merchants, along with Aleuts forcibly relocated from the western region of the Aleutian archipelago. The purpose of this study is to characterize the genetic variation of Bering Island inhabitants for autosomal, mitochondrial, and Y-chromosome DNA and classic genetic markers and to investigate the genetic impact of the 19th-century founding and subsequent demographic events on this heterogeneous community. Our results show a loss of diversity among maternal lineages in the Bering Aleut population, with fixation of mtDNA haplogroup D, as revealed by median-joining network analysis and mismatch differences. Conversely, paternal haplotypes exhibit an increase in diversity and the presence of a substantial number of non-Native lineages. Admixture results, based on autosomal STR data, indicate that parental contributions to the mixed Aleut population of Bering are approximately 60% Aleut and 40% Russian. Classic genetic markers show affinities between the Bering Island Aleuts and the other historically founded Aleut communities of St. Paul and St. George in the Pribilof Islands, Alaska. This study demonstrates that the opposing evolutionary forces of genetic drift and gene flow acted on the maternal and paternal lineages, respectively, to shape the genetic structure of the present-day inhabitants of Bering Island.

Genetic diversity and phylogeographic analysis of human herpesvirus type 8 (HHV-8) in two distant regions of Argentina: Association with the genetic ancestry of the population.

BACKGROUND: The genetic diversity of persistent infectious agents, such as HHV-8, correlates closely with the migration of modern humans out of East Africa which makes them useful to trace human migrations. However, there is scarce data about the evolutionary history of HHV-8 particularly in multiethnic Latin American populations. OBJECTIVES: The aims of this study were to characterize the genetic diversity and the phylogeography of HHV-8 in two distant geographic regions of Argentina, and to establish potential associations with pathogenic conditions and the genetic ancestry of the population. STUDY DESIGN: A total of 101 HIV-1 infected subjects, 93 Kaposi's Sarcoma (KS) patients and 411 blood donors were recruited in the metropolitan (MET) and north-western regions of Argentina (NWA). HHV-8 DNA was detected by ORF-26 PCR in whole blood, saliva and FFPE tissues. Then, ORF-26 and ORF-K1 were analyzed for subtype assignment. Mitochondrial DNA and Y chromosome haplogroups, as well as autosomal ancestry markers were evaluated in samples in which subtypes could be assigned. Phylogeographic analysis was performed in the ORF-K1 sequences from this study combined with 388 GenBank sequences. RESULTS: HHV-8 was detected in 50.7%, 59.2% and 8% of samples from HIV-1 infected subjects, KS patients and blood donors, respectively. ORF-K1 phylogenetic analyses showed that subtypes A (A1-A5), B1, C (C1-C3) and F were present in 46.9%, 6.25%, 43.75% and 3.1% of cases, respectively. Analyses of ORF-26 fragment revealed that 81.95% of strains were subtypes A/C followed by J, B2, R, and K. The prevalence of subtype J was more commonly observed among KS patients when compared to the other groups. Among KS patients, subtype A/C was more commonly detected in MET whereas subtype J was the most frequent in NWA. Subtypes A/C was significantly associated with Native American maternal haplogroups (p = 0.004), whereas subtype J was related to non-Native American haplogroups (p < 0.0001). Sub-Saharan Africa, Europe and Latin America were the most probable locations from where HHV-8 was introduced to Argentina. CONCLUSIONS: These results give evidence of the geographic circulation of HHV-8 in Argentina, suggest the association of ORF-26 subtype J with KS development and provide new insights about its relationship with ancient and modern human migrations and identify the possible origins of this virus in Argentina.

Bi-directional Recurrent Neural Network Models for Geographic Location Extraction in Biomedical Literature.

Phylogeography research involving virus spread and tree reconstruction relies on accurate geographic locations of infected hosts. Insufficient level of geographic information in nucleotide sequence repositories such as GenBank motivates the use of natural language processing methods for extracting geographic location names (toponyms) in the scientific article associated with the sequence, and disambiguating the locations to their co-ordinates. In this paper, we present an extensive study of multiple recurrent neural network architectures for the task of extracting geographic locations and their effective contribution to the disambiguation task using population heuristics. The methods presented in this paper achieve a strict detection F1 score of 0.94, disambiguation accuracy of 91% and an overall resolution F1 score of 0.88 that are significantly higher than previously developed methods, improving our capability to find the location of infected hosts and enrich metadata information.

Identifying suitable habitat and corridors for Indian Grey Wolf (Canis lupus pallipes) in Chotta Nagpur Plateau and Lower Gangetic Planes: A species with differential management needs.

Different Biogeographic provinces and environmental factors are known to influence the dispersibility of long-ranging carnivores over the landscape. However, lack of empirical data on long-ranging carnivores may lead to erroneous decisions in formulating management plans. The Indian Grey wolf (Canis lupus pallipes) is known to be distributed in the vast areas of the Indian subcontinent. However, the actual population estimates are available only for Gujarat, Karnataka, Rajasthan and Bihar. Whereas, its distribution, population and habitat ecology is poorly known from the eastern region. Hence, this article aimed to evaluate the habitat suitability along with landscape connectivity for the species over the two major biogeographic provinces of India, i.e., Lower Gangetic Plains (7b) and Chhota Nagpur Plateau (6b). The present model with significantly higher Area under the curve (AUC) value of 0.981, indicates its accuracy in predicting the suitable habitats and identifying biological corridors by using environmental, topological and anthropogenic variables. Precipitation of the driest quarter and the precipitation of seasonality were the two best performing variables in our model, capable of explaining about 26% and 22.4% variation in the data respectively. Out of the total area i.e. 4,16,665 Km2, about 18,237 Km2 (4.37%) was found to be highly suitable area and about 3,16,803 Km2 (76.03%) areas as least suitable. The corridor analysis indicated that the habitat connectivity was highest in the border line area of the two biotic provinces located in the south-eastern zone via districts of Purba Singhbhum and Paschim Singhbhum of Jharkhand state and Bankura and West Midnapore districts of West Bengal state. Among the Protected Areas (PAs), natural corridors exist connecting the Simlipal National Park (NP)-Satkosia Wildlife Santuray (WLS), Dalma ranges of Chotta Nagpur plateau along with Badrama WLS, Khulasuni WLS and Debrigarh WLS. Differential management through landscape level planning may be helpful in securing the future of the species in the landscape.

Phylogeography, genetic diversity, and population structure of Nile crocodile populations at the fringes of the southern African distribution.

Nile crocodiles are apex predators widely distributed in sub-Saharan Africa that have been viewed and managed as a single species. A complex picture of broad and fine-scale phylogeographic patterns that includes the recognition of two species (Crocodylus niloticus and Crocodylus suchus), and the structuring of populations according to river basins has started to emerge. However, previous studies surveyed a limited number of samples and geographical regions, and large areas of the continent remained unstudied. This work aimed at a fine scale portrait of Nile crocodile populations at the fringes of their geographic distribution in southern Africa. Wild and captive individuals were sampled across four major river systems (Okavango, Lower Kunene, Lower Shire and Limpopo) and the KwaZulu-Natal region. A multi-marker approach was used to infer phylogeographic and genetic diversity patterns, including new and public mitochondrial data, and a panel of 11 nuclear microsatellites. All individuals belonged to a phylogenetic clade previously associated with the C. niloticus species, thus suggesting the absence of C. suchus in southern Africa. The distribution of mitochondrial haplotypes indicated ancestral genetic connectivity across large areas, with loss of diversity along the north-south axis. Genetic variation partitioned the populations primarily into western and eastern regions of southern Africa, and secondarily into the major river systems. Populations were partitioned into five main groups corresponding to the Lower Kunene, the Okavango, the Lower Shire, and the Limpopo rivers, and the KwaZulu-Natal coastal region. All groups show evidence of recent bottlenecks and small effective population sizes. Long-term genetic diversity is likely to be compromised, raising conservation concern. These results emphasize the need for local genetic assessment of wild populations of Nile crocodiles to inform strategies for management of the species in southern Africa.

Phylogeny, biogeography and methodology: a meta-analytic perspective on heterogeneity in adult marine turtle survival rates.

Comparative syntheses of key demographic parameters are critical not only for identifying data gaps, but also for evaluating sources of heterogeneity among estimates. Because demographic studies frequently exhibit heterogeneity, evaluating sources of heterogeneity among estimates can inform biological patterns and conservation actions more broadly. To better understand adult survival in marine turtles and avoid drawing inaccurate conclusions from current estimates, we conducted a comprehensive meta-analysis to test how heterogeneity among estimates was partitioned among phylogenetic, biogeographic and methodological factors. Fifty-nine studies from five marine turtle species met the minimum selection criteria for inclusion in our meta-analysis. Heterogeneity among survival estimates was first partitioned between differences in ocean basin (Indo-Pacific versus Atlantic), then by differences among family/tribe within the Indo-Pacific (Chelonini versus Carettini and Dermochelidae). However, apparent differences attributed to biogeography (ocean basin effect) and phylogeny (family/tribe effect) were highly correlated with methodological differences in tag type, model type, habitat type and study duration, thereby confounding biological interpretations and complicating efforts to use many current survival estimates in population assessments. Our results highlight the importance of evaluating sources of heterogeneity when interpreting patterns among similar demographic studies and directly inform efforts to identify research priorities for marine turtles globally.

Changing patterns of human migrations shaped the global population structure of Mycobacterium tuberculosis in France.

Mycobacterium tuberculosis (Mtb) exhibits a structured phylogeographic distribution worldwide linked with human migrations. We sought to infer how the interactions between distinct human populations shape the global population structure of Mtb on a regional scale. We applied the recently described timescaled haplotypic density (THD) technique on 638 minisatellite-based Mtb genotypes from French tuberculosis patients. THD with a long-term (200 y) timescale indicated that Mtb population in France had been mostly influenced by interactions with Eastern and Southern Europe and, to a lesser extent, Northern and Middle Africa, consistent with historical migrations favored by geographic proximity or commercial exchanges with former French colonies. Restricting the timescale to 20 y, THD identified a sustained influence of Northern Africa, but not Europe where tuberculosis incidence decreased sharply. Evolving interactions between human populations, thus, measurably influence the local population structure of Mtb. Relevant information on such interactions can be inferred using THD from Mtb genotypes.

An Efficient Optimization Method for Solving Unsupervised Data Classification Problems.

Unsupervised data classification (or clustering) analysis is one of the most useful tools and a descriptive task in data mining that seeks to classify homogeneous groups of objects based on similarity and is used in many medical disciplines and various applications. In general, there is no single algorithm that is suitable for all types of data, conditions, and applications. Each algorithm has its own advantages, limitations, and deficiencies. Hence, research for novel and effective approaches for unsupervised data classification is still active. In this paper a heuristic algorithm, Biogeography-Based Optimization (BBO) algorithm, was adapted for data clustering problems by modifying the main operators of BBO algorithm, which is inspired from the natural biogeography distribution of different species. Similar to other population-based algorithms, BBO algorithm starts with an initial population of candidate solutions to an optimization problem and an objective function that is calculated for them. To evaluate the performance of the proposed algorithm assessment was carried on six medical and real life datasets and was compared with eight well known and recent unsupervised data classification algorithms. Numerical results demonstrate that the proposed evolutionary optimization algorithm is efficient for unsupervised data classification.

The spatiotemporal expansion of human rabies and its probable explanation in mainland China, 2004-2013.

BACKGROUND: Human rabies is a significant public health concern in mainland China. However, the neglect of rabies expansion and scarce analyses of the dynamics have made the spatiotemporal spread pattern of human rabies and its determinants being poorly understood. METHODS: We collected geographic locations and timeline of reported human rabies cases, rabies sequences and socioeconomic variables for the years 2004-2013, and integrated multidisciplinary approaches, including epidemiological characterization, hotspots identification, risk factors analysis and phylogeographic inference, to explore the spread pattern of human rabies in mainland China during the last decade. RESULTS: The results show that human rabies distribution and hotspots were expanding from southeastern regions to north or west regions, which could be associated with the evolution of the virus, especially the clade I-G. A Panel Poisson Regression analysis reveals that human rabies incidences had significant correlation with the education level, GDP per capita, temperature at one-month lag and canine rabies outbreak at two-month lag. CONCLUSIONS: The reduction in the overall human rabies incidence was accompanied by a westward and northward expansion of the circulating region in mainland China. Higher risk of human rabies was associated with lower level of education and economic status. New clades of rabies, especial Clade I-G, played an important role in recent spread. Our findings provide valuable information for rabies control and prevention in the future.