Stochastic Search Variable Selection Research Articles

Sequential Learning and Economic Benefits from Dynamic Term Structure Models

We explore the statistical and economic importance of restrictions on the dynamics of risk compensation from the perspective of a real-time Bayesian learner who predicts bond excess returns using dynamic term structure models (DTSMs). The question on whether potential statistical predictability offered by such models can generate economically significant portfolio benefits out-of-sample is revisited while imposing restrictions on their risk premia parameters. To address this question, we propose a methodological framework that successfully handles sequential model search and parameter estimation over the restriction space in real time, allowing investors to revise their beliefs when new information arrives, thus informing their asset allocation and maximizing their expected utility. Empirical results reinforce the argument of sparsity in the market price of risk specification since we find strong evidence of out-of-sample predictability only for those models that allow for level risk to be priced and, additionally, only one or two of these risk premia parameters to be different than zero. Most importantly, such statistical evidence is turned into economically significant utility gains, across prediction horizons, different time periods and portfolio specifications. In addition to identifying successful DTSMs, the sequential version of the stochastic search variable selection scheme developed can be applied on its own and offer useful diagnostics monitoring key quantities over time. Connections with predictive regressions are also provided. This paper was accepted by Kay Giesecke, finance. Funding: T. Dubiel-Teleszynski acknowledges the support of the Economic and Social Research Council. Supplemental Material: The data files and online appendices are available at https://doi.org/10.1287/mnsc.2023.4801 .

Liver fibrosis is associated with left ventricular remodeling: insights into the liver-heart axis

Abstract Background Non-alcoholic fatty liver disease (NAFLD), common in type 2 diabetes, has emerged as an important and independent risk factor for adverse cardiac remodeling. Fibrosis in the liver could be a key factor in this relationship. Despite the fact that NAFLD is a long-term progressive disease, imaging studies on this topic are mainly limited to selected patients from tertiary centres. Purpose To investigate the relationship between liver fat, liver fibrosis and cardiac remodeling in the general population. We also sought to explore biomarkers of cardiovascular inflammation and fibrosis, for possible mechanistic pathways in these relationships. Methods We included 46 participants with type 2 diabetes and 46 age and sex-matched controls (overall 35%/65% F/M, mean age 59.5 ± 4.6), all randomly sampled from the general population. Participants underwent careful clinical phenotyping, including blood sampling, echocardiography, and magnetic resonance imaging (MRI) at 1.5 T. The MRI protocol included sequences for quantification of liver proton density fat fraction (spectroscopy), liver fibrosis (stiffness from elastography at 56 Hz), left ventricular (LV) structure, function, and tissue characteristics (native T1 mapping). We also analyzed ∼290 circulating cardiovascular biomarkers. A Bayesian statistical approach (‘Stochastic Search Variable Selection’) was used to identify the biomarkers most strongly associated with cardiac remodeling. Results For liver stiffness, the mean value was 2.1 ± 0.4 kPa and the highest value was 2.9 kPa. The median liver fat value was 5.7% (IQR 9.9%) and 53 participants (58%) displayed hepatic steatosis (liver fat > 5%). LV concentricity increased across quartiles of liver stiffness (P=0.0014 for ANOVA), while LV concentricity and diastolic dysfunction increased across quartiles of liver fat (P=0.0023 and P=0.030 for ANOVA, respectively). Myocardial T1 displayed no relationships to neither liver fat nor liver stiffness. In a multiple linear regression, liver stiffness was positively associated with LV concentricity (Beta=0.26, P=0.0053), independently of diabetes and liver fat. This association was attenuated (Beta=0.17, P=0.077) when adjusting for circulating levels of IL-1 receptor type 2, the biomarker with the strongest association to LV concentricity in the Bayesian statistical analysis. Conclusion In our well-characterized subjects from the general population, we found that a majority had hepatic steatosis but overall rather low liver stiffness values. Despite this, liver stiffness was related to increased LV concentricity independently of liver fat and diabetes. This association was attenuated when also adjusting for IL-1 receptor type 2. Our results suggest a link between liver fibrosis and cardiac remodeling also in the early stages of fatty liver disease, possibly involving the IL-1 signaling pathway. Future studies are needed to further establish causality and potential mechanisms.Multiple linear regression analysesLiver magnetic resonance elastogram

Application of statistical machine learning in biomarker selection

In the recent JAVELIN Bladder 100 phase 3 trial, avelumab plus best supportive care significantly prolonged overall survival relative to best supportive care alone as first-line maintenance therapy following first-line platinum-based chemotherapy in patients with advanced urothelial cancer (aUC). Discovering biomarkers using genomic profiling to understand potential patient heterogeneity is essential to help improve patient care with precision medicine. For the JAVELIN Bladder 100 trial, it is unclear which variable selection methods can most reliably identify biomarkers to inform patient care because the dataset is characterized by high collinearity and low signal. The aim of this paper was to evaluate available selection methods and their ability to discover prognostic and predictive biomarkers in patients with aUC receiving first-line maintenance therapy. A simulation study evaluated the performance of popular variable selection approaches for high-dimensional data including penalized regression models, random survival forests, and Bayesian variable selection methods. For Bayesian variable selection methods, a modified Bayesian Information Criterion (BIC) thresholding rule was proposed in addition to the traditional BIC thresholding rule. These methods were applied to the JAVELIN Bladder 100 dataset to investigate potential biomarkers associated with survival benefit. Results from the simulations demonstrated the strengths and limitations of the different methods. The variable selection methods demonstrated low false discovery rates under different conditions. However, their performance declined in the presence of high collinearity. Using the JAVELIN Bladder 100 data, we identified some potentially significant biomarkers across multiple models. Several lasso-related methods were able to identify potentially biologically meaningful variables in the trial. Some variable selection methods (such as stochastic search variable selection and random survival forest) may not be well suited to this type of data due to the presence of extreme collinearity and low signal. Future research should explore novel variable selection methods that may be more suitable for identifying prognostic and predictive biomarkers in this population.Trial registration: ClinicalTrials.gov Identifier: NCT02603432.

Inconsistency identification in network meta-analysis via stochastic search variable selection.

The reliability of the results of network meta-analysis (NMA) lies in the plausibility of the key assumption of transitivity. This assumption implies that the effect modifiers' distribution is similar across treatment comparisons. Transitivity is statistically manifested through the consistency assumption which suggests that direct and indirect evidence are in agreement. Several methods have been suggested to evaluate consistency. A popular approach suggests adding inconsistency factors to the NMA model. We follow a different direction by describing each inconsistency factor with a candidate covariate whose choice relies on variable selection techniques. Our proposed method, stochastic search inconsistency factor selection (SSIFS), evaluates the consistency assumption both locally and globally, by applying the stochastic search variable selection method to determine whether the inconsistency factors should be included in the model. The posterior inclusion probability of each inconsistency factor quantifies how likely is a specific comparison to be inconsistent. We use posterior model odds or the median probability model to decide on the importance of inconsistency factors. Differences between direct and indirect evidence can be incorporated into the inconsistency detection process. A key point of our proposed approach is the construction of a reasonable "informative" prior concerning network consistency. The prior is based on the elicitation of information derived historical data from 201 published network meta-analyses. The performance of our proposed method is evaluated in two published network meta-analyses. The proposed methodology is publicly available in an R package called ssifs, published on CRAN and developed and maintained by the authors of this work.

Separating Behaviours and Adverse Consequences in the Problem Gambling Severity Index (PGSI): A Confirmatory Factor Analysis and Rasch Analysis.

The Problem Gambling Severity Index (PGSI) is considered the "gold standard" for measuring problem gambling. The PGSI provides a single score summed across nine items. The nine items of the PGSI comprise two subdomains: problematic behaviours and adverse consequences. The aim of the present study was to compare evidence of a one-factor structure to evidence of a two-factor structure representing the two subdomains. With a sample of 1,251 bettors, we conducted confirmatory factor analyses and Rasch analyses to assess evidence supporting the one-factor and the two-factor structures. In addition, stochastic search variable selection was conducted with the total PGSI score, PGSI behaviour score, and PGSI adverse consequences score as separate outcomes to examine whether information is lost when collapsing the two subdomains into a single factor. Overall, there was stronger support of a two-factor structure than a one-factor structure. However, the two-factors were highly correlated with one another and shared most predictors except for one. We recommend continued use of the one-factor structure of the PGSI unless one aims to better understand the relationship between problematic behaviours and adverse consequences.

Bayesian Subset Selection of Seasonal Autoregressive Models

Seasonal autoregressive (SAR) models have many applications in different fields, such as economics and finance. It is well known in the literature that these models are nonlinear in their coefficients and that their Bayesian analysis is complicated. Accordingly, choosing the best subset of these models is a challenging task. Therefore, in this paper, we tackled this problem by introducing a Bayesian method for selecting the most promising subset of the SAR models. In particular, we introduced latent variables for the SAR model lags, assumed model errors to be normally distributed, and adopted and modified the stochastic search variable selection (SSVS) procedure for the SAR models. Thus, we derived full conditional posterior distributions of the SAR model parameters in the closed form, and we then introduced the Gibbs sampler, along with SSVS, to present an efficient algorithm for the Bayesian subset selection of the SAR models. In this work, we employed mixture–normal, inverse gamma, and Bernoulli priors for the SAR model coefficients, variance, and latent variables, respectively. Moreover, we introduced a simulation study and a real-world application to evaluate the accuracy of the proposed algorithm.

Comparing Bayesian Variable Selection to Lasso Approaches for Applications in Psychology

In the current paper, we review existing tools for solving variable selection problems in psychology. Modern regularization methods such as lasso regression have recently been introduced in the field and are incorporated into popular methodologies, such as network analysis. However, several recognized limitations of lasso regularization may limit its suitability for psychological research. In this paper, we compare the properties of lasso approaches used for variable selection to Bayesian variable selection approaches. In particular we highlight advantages of stochastic search variable selection (SSVS), that make it well suited for variable selection applications in psychology. We demonstrate these advantages and contrast SSVS with lasso type penalization in an application to predict depression symptoms in a large sample and an accompanying simulation study. We investigate the effects of sample size, effect size, and patterns of correlation among predictors on rates of correct and false inclusion and bias in the estimates. SSVS as investigated here is reasonably computationally efficient and powerful to detect moderate effects in small sample sizes (or small effects in moderate sample sizes), while protecting against false inclusion and without over-penalizing true effects. We recommend SSVS as a flexible framework that is well-suited for the field, discuss limitations, and suggest directions for future development.

Global and local components of output gaps

This paper proposes a multi-level dynamic factor model to identify common components in output gap estimates. We pool multiple estimates for 157 countries and decompose them into one global, eight regional, and 157 country-specific cycles. Our approach easily deals with mixed frequencies, ragged edges, and discontinuities in the underlying output gap estimates. To restrict the parameter space in the Bayesian state space model, we apply a stochastic search variable selection approach and base the prior inclusion probabilities on spatial information. Our results suggest that the global and the regional cycles explain a substantial proportion of the output gaps. On average, 18% of a country’s output gap is attributable to the global cycle, 24% to the regional cycle, and 58% to the local cycle.

Factors Associated with Antiretroviral Therapy Adherence Among a Community-Based Sample of Sexual Minority Older Adults with HIV.

Older sexual minorities (e.g., gay, bisexual) living with HIV are at risk for poor HIV outcomes due to their frequent experience with both psychosocial challenges and structural barriers to care. This study utilized a stochastic search variable selection (SVSS) approach to explore potential psychosocial and structural factors associated with HIV-related health outcomes among a community-based sample of older sexual minorities (N = 150) in South Florida, an U.S. HIV-epidemic epicenter. After SVSS, a forward entry regression approach suggested unstable housing, illicit substance use, current nicotine use, and depression were all associated with poorer ART adherence among older sexual minority adults living with HIV. No associations between potential correlates and biological measures of HIV disease severity were observed. Findings highlight a need to focus on multiple levels of intervention that target a combination of psychosocial and structural factors to improve HIV-care outcomes among older sexual minorities and achieve Ending the HIV Epidemic goals.

Bayesian Variable Selection in Generalized Extreme Value Regression: Modeling Annual Maximum Temperature

In many applications, interest focuses on assessing relationships between covariates and the extremes of the distribution of a continuous response. For example, in climate studies, a usual approach to assess climate change has been based on the analysis of annual maximum data. Using the generalized extreme value (GEV) distribution, we can model trends in the annual maximum temperature using the high number of available atmospheric covariates. However, there is typically uncertainty in which of the many candidate covariates should be included. Bayesian methods for variable selection are very useful to identify important covariates. However, such methods are currently very limited for moderately high dimensional variable selection in GEV regression. We propose a Bayesian method for variable selection based on a stochastic search variable selection (SSVS) algorithm proposed for posterior computation. The method is applied to the selection of atmospheric covariates in annual maximum temperature series in three Spanish stations.

Bayesian outcome selection modeling.

In psychiatric and social epidemiology studies, it is common to measure multiple different outcomes using a comprehensive battery of tests thought to be related to an underlying construct of interest. In the research that motivates our work, researchers wanted to assess the impact of in utero alcohol exposure on child cognition and neuropsychological development, which are evaluated using a range of different psychometric tests. Statistical analysis of the resulting multiple outcomes data can be challenging, because the outcomes measured on the same individual are not independent. Moreover, it is unclear, a priori, which outcomes are impacted by the exposure under study. While researchers will typically have some hypotheses about which outcomes are important, a framework is needed to help identify outcomes that are sensitive to the exposure and to quantify the associated treatment or exposure effects of interest. We propose such a framework using a modification of stochastic search variable selection, a popular Bayesian variable selection model and use it to quantify an overall effect of the exposure on the affected outcomes. The performance of the method is investigated empirically and an illustration is given through application using data from our motivating study.

Engaging Latino sexual minority men in PrEP and behavioral health care: multilevel barriers, facilitators, and potential implementation strategies.

Latino sexual minority men (LSMM) are affected by HIV and behavioral health disparities. Evidence-based HIV-prevention and behavioral health (BH) services are not sufficiently scaled up to LSMM. The current study identified multilevel barriers and facilitators to LSMM's use of HIV-prevention and BH services. LSMM (N = 290) in South Florida, a US HIV epicenter, completed a battery of measures potentially associated with pre-exposure prophylaxis (PrEP) and BH treatment use. Stochastic search variable selection (SSVS) followed by multiple linear regression analyses identified variables associated with engagement in PrEP and BH treatment. Multilevel determinants of PrEP and BH treatment engagement were identified, with most identified determinants being at the relational level (e.g., stigma, discrimination based on income and immigration status, personal recommendation for treatment). Individual (e.g., knowledge, self-efficacy) and structural (e.g., financial stress) determinants were also identified. Accordingly, modifiable leverage points to enhance the reach of PrEP and BH treatment to LSMM include educating and enhancing the perceived relevance of services, de-stigmatizing and normalizing via peer examples, bolstering self-efficacy, and building trust.

Factors Associated with Post-Exposure Prophylaxis Awareness Among Latino Sexual Minority Men in South Florida.

Despite their efficacy, biomedical HIV prevention tools such as post-exposure prophylaxis (PEP) and pre-exposure prophylaxis (PrEP) have been insufficiently scaled up and disseminated, especially among marginalized subgroups that face substantial HIV disparities. Given the minimal literature available on PEP among Latino sexual minority men (LSMM), this cross-sectional secondary analysis explored factors associated with PEP awareness among a group of LSMM living in South Florida, a US HIV epicenter. The parent study examined patterns of engagement in PrEP and behavioral health treatment services among LSMM (N = 290). The current secondary analysis (N = 243) identified factors associated with PEP awareness using three methods: stochastic search variable selection, participatory data science, and literature review-before being modeled using linear regression. Most participants (67.5%) reported having little to no awareness about PEP before initiating our study. Simple linear regression models suggested that higher PrEP knowledge (B = 0.17, SE = 0.02, p < 0.001), HIV knowledge (B = 0.15, SE = 0.04, p < 0.001), PrEP self-efficacy (B = 0.37, SE = 0.13, p < 0.05), and high perceived community norms for HIV testing (B = 0.29, SE = 0.14, p < 0.05) were each associated with LSMM's greater PEP awareness, while identity affirmation was associated with less PEP awareness (B = -0.13, SE = 0.05, p < 0.01). Results suggest the utility of our three-pronged variable selection approach and address gaps in PEP awareness and use among LSMM living in a US HIV epicenter to support Ending the HIV Epidemic goals.

Proteomics data analysis using multiple statistical approaches identified proteins and metabolic networks associated with sucrose accumulation in sugarcane

BackgroundSugarcane is the most important sugar crop, contributing > 80% of global sugar production. High sucrose content is a key target of sugarcane breeding, yet sucrose improvement in sugarcane remains extremely slow for decades. Molecular breeding has the potential to break through the genetic bottleneck of sucrose improvement. Dissecting the molecular mechanism(s) and identifying the key genetic elements controlling sucrose accumulation will accelerate sucrose improvement by molecular breeding. In our previous work, a proteomics dataset based on 12 independent samples from high- and low-sugar genotypes treated with ethephon or water was established. However, in that study, employing conventional analysis, only 25 proteins involved in sugar metabolism were identified .ResultsIn this work, the proteomics dataset used in our previous study was reanalyzed by three different statistical approaches, which include a logistic marginal regression, a penalized multiple logistic regression named Elastic net, as well as a Bayesian multiple logistic regression method named Stochastic search variable selection (SSVS) to identify more sugar metabolism-associated proteins. A total of 507 differentially abundant proteins (DAPs) were identified from this dataset, with 5 of them were validated by western blot. Among the DAPs, 49 proteins were found to participate in sugar metabolism-related processes including photosynthesis, carbon fixation as well as carbon, amino sugar, nucleotide sugar, starch and sucrose metabolism. Based on our studies, a putative network of key proteins regulating sucrose accumulation in sugarcane is proposed, with glucose-6-phosphate isomerase, 2-phospho-D-glycerate hydrolyase, malate dehydrogenase and phospho-glycerate kinase, as hub proteins.ConclusionsThe sugar metabolism-related proteins identified in this work are potential candidates for sucrose improvement by molecular breeding. Further, this work provides an alternative solution for omics data processing.

A Bayesian method for identifying associations between response variables and bacterial community composition.

Determining associations between intestinal bacteria and continuously measured physiological outcomes is important for understanding the bacteria-host relationship but is not straightforward since abundance data (compositional data) are not normally distributed. To address this issue, we developed a fully Bayesian linear regression model (BRACoD; Bayesian Regression Analysis of Compositional Data) with physiological measurements (continuous data) as a function of a matrix of relative bacterial abundances. Bacteria can be classified as operational taxonomic units or by taxonomy (genus, family, etc.). Bacteria associated with the physiological measurement were identified using a Bayesian variable selection method: Stochastic Search Variable Selection. The output is a list of inclusion probabilities () and coefficients that indicate the strength of the association () for each bacterial taxa. Tests with simulated communities showed that adopting a cut point value of ≥ 0.3 for identifying included bacteria optimized the true positive rate (TPR) while maintaining a false positive rate (FPR) of ≤ 5%. At this point, the chances of identifying non-contributing bacteria were low and all well-established contributors were included. Comparison with other methods showed that BRACoD (at ≥ 0.3) had higher precision and a higher TPR than a commonly used center log transformed LASSO procedure (clr-LASSO) as well as higher TPR than an off-the-shelf Spike and Slab method after center log transformation (clr-SS). BRACoD was also less likely to include non-contributing bacteria that merely correlate with contributing bacteria. Analysis of a rat microbiome experiment identified 47 operational taxonomic units that contributed to fecal butyrate levels. Of these, 31 were positively and 16 negatively associated with butyrate. Consistent with their known role in butyrate metabolism, most of these fell within the Lachnospiraceae and Ruminococcaceae. We conclude that BRACoD provides a more precise and accurate method for determining bacteria associated with a continuous physiological outcome compared to clr-LASSO. It is more sensitive than a generalized clr-SS algorithm, although it has a higher FPR. Its ability to distinguish genuine contributors from correlated bacteria makes it better suited to discriminating bacteria that directly contribute to an outcome. The algorithm corrects for the distortions arising from compositional data making it appropriate for analysis of microbiome data.

Post-Cold War civil conflict and the role of history and religion: A stochastic search variable selection approach

Despite colossal economic and human losses caused by conflict and violence, designing effective policies to avoid conflict remains challenging. While the literature has proposed a voluminous set of candidate predictors, their robustness is questionable and model uncertainty masks the true drivers of conflicts and wars. Considering a comprehensive set of 34 potential determinants in 175 post-Cold-War countries, we employ stochastic search variable selection (SSVS) to sort through all 234 possible models to address model uncertainty. We find past conflict constitutes the most powerful predictor of current conflict: Path dependency matters. Also, larger shares of Jewish, Muslim, or Christian citizens are associated with increased conflict, while economic and political factors remain less relevant than colonial origin and religion. Our results help future researchers and policymakers by inching towards causality and providing a standard set of covariates that need to be accounted for in designing any relevant policies.

Small open economies and external shocks: an application of Bayesian global vector autoregression model.

This study assesses the impact of external shocks on select small open economies (SOEs) using the Bayesian variant of the global vector autoregression model with time varying parameters and stochastic volatility. We account for the curse of dimensionality in the multi-country VAR system by implementing three different priors in the estimation of the parameters of the model: the Minnesota (M-N) prior of Doan–Litterman et al. (1984; Litterman 1986); the Normal-Gamma (N-G) prior of Park and Casella (Bayesian Anal 1:515–533, 2008); and the Stochastic Search Variable Selection (SSVS) prior of George and McCulloch (1995) as extended by Koop and Korobilis (2010, 2013). From our simulation results, we found that global economies of the USA, Western Europe and China are the major drivers of cyclical fluctuation in the SOEs. However, in spite of the perceived superior influence of China on the SOEs GDPs’ response to external shocks, we found no evidence to conclude that the influence is significantly greater than those exerted by the United States or Europe on the bloc’s economies.

Forecasting with Bayesian vector autoregressive models: comparison of direct and iterated multistep methods

PurposeThe paper compares multi-period forecasting performances by direct and iterated method using Bayesian vector autoregressive (VAR) models.Design/methodology/approachThe paper adopts Bayesian VAR models with three different priors – independent Normal-Wishart prior, the Minnesota prior and the stochastic search variable selection (SSVS). Monte Carlo simulations are conducted to compare forecasting performances. An empirical study using US macroeconomic data are shown as an illustration.FindingsIn theory direct forecasts are more efficient asymptotically and more robust to model misspecification than iterated forecasts, and iterated forecasts tend to bias but more efficient if the one-period ahead model is correctly specified. From the results of the Monte Carlo simulations, iterated forecasts tend to outperform direct forecasts, particularly with longer lag model and with longer forecast horizons. Implementing SSVS prior generally improves forecasting performance over unrestricted VAR model for either nonstationary or stationary data.Originality/valueThe paper finds that iterated forecasts using model with the SSVS prior generally best outperform, suggesting that the SSVS restrictions on insignificant parameters alleviates over-parameterized problem of VAR in one-step ahead forecast and thus offers an appreciable improvement in forecast performance of iterated forecasts.

The Genomic Evolution and the Transmission Dynamics of H6N2 Avian Influenza A Viruses in Southern China

In China, the broad prevalence of H6 subtype influenza viruses, increasingly detected in aquatic birds, promotes their exchange materials with other highly pathogenic human-infecting H5N1, H5N6, and H7N9 influenza viruses. Strikingly, some H6 subtype viruses can infect pigs, dogs, and humans, posing risks to public health. In this study, 9 H6N2 viruses recovered from waterfowl species in the Guangdong province of China in 2018 were isolated and sequenced. Phylogenetic analysis revealed that the genome sequences of these H6N2 viruses belonged to Group I, except for the NP gene in Group III. Coalescent analyses demonstrated that the reassortment of NA and NS genes have occurred in two independent clusters, suggesting H6 subtype viruses had been undergoing a complex reassortant. To examine the evolutionary dynamics and the dissemination of the H6 subtype viruses, a Bayesian stochastic search variable selection was performed for results showing higher viral migration rates between closer provinces, including Guangdong, Jiangxi, Guangxi, and Fujian. Notably, the transmission routes of the H6 subtype viruses were concentrated in Jiangxi Province, the most frequent location for input and output transmission and a region containing Poyang Lake, a well-known wintering site for migration birds. We also found that the aquatic birds, especially ducks, were the most common input source of the viral transmission. In addition, we also found that eight positively selected amino acid sites were identified in HA protein. Given their continuous dissemination and the broad prevalence of the H6 subtype influenza viruses, continued surveillance is warranted in the future.

Dating the origin and dispersal of global hepatitis B virus genotype C in humans.

Hepatitis B virus genotype C (HBV/C) is one of the most prevalent HBV strains worldwide, especially in the Western Pacific and the South-East Asia. However, the origin and evolutionary timescale of HBV/C remains largely unresolved. We analyzed the evolutionary rate and molecular clock phylogeny of 101 full-genome HBV/C sequences sampled globally using a Bayesian Markov Chain Monte Carlo (MCMC) approach. We inferred the spatiotemporal dynamics of the HBV/C worldwide by the Bayesian Stochastic Search Variable Selection (BSSVS). We found that the estimated mean evolution rate of the HBV/C genotype full-genome was 4.32 × 10-5 subs/site/year (95% highest posterior density 3.02 × 10-6 - 8.97 × 10-5). Phylogeographic reconstruction was able to identify a single location for the origin of the global HBV/C in Australia around A.D. 715. The subgenotype C4 diverged earliest and mainly circulated in Australia, C1 mainly in Southeast Asia, C2 mainly in East Asia and C3 in Remote Oceania. The effective number of HBV infection presented a rapid exponential increase between the 1760s and 1860s followed by a maintained high level until now. Our study, for the first time, provides an estimated timescale for the HBV/C epidemic, and brings new insight to the dispersal of HBV/C in humans globally. Based on the continuous presence of a highly effective viral population, this study provides further evidence of the challenge from a population-based molecular level to eliminate HBV by 2030, and calls for a concerted effort from policy makers, health providers, and society in the globalized world.