Generic reversible jump MCMC using graphical models

Lunn, David J.; Best, Nicky; Whittaker, John C.

doi:10.1007/s11222-008-9100-0

Cited by 86 publications

(96 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, reversible jump methods were implemented for a generic class of problems [52], but the main focus of this work became applications in genetic epidemiology [53] and pharmacokinetics [54]. In addition, the differential equation solving capabilities were strengthened to accommodate more complex models in pharmacokinetics and make them more applicable to infectious disease modelling.…”

Section: Divergence Of Established and Experimental Versions In 2004mentioning

confidence: 99%

The BUGS project: Evolution, critique and future directions

Lunn

Spiegelhalter

Thomas

et al. 2009

Statistics in Medicine

Self Cite

1,724

1,323

View full text Add to dashboard Cite

SUMMARYBUGS is a software package for Bayesian inference using Gibbs sampling. The software has been instrumental in raising awareness of Bayesian modelling among both academic and commercial communities internationally, and has enjoyed considerable success over its 20-year life span. Despite this, the software has a number of shortcomings and a principal aim of this paper is to provide a balanced critical appraisal, in particular highlighting how various ideas have led to unprecedented flexibility while at the same time producing negative side effects. We also present a historical overview of the BUGS project and some future perspectives.

show abstract

Section: Divergence Of Established and Experimental Versions In 2004mentioning

confidence: 99%

The BUGS project: Evolution, critique and future directions

Lunn

Spiegelhalter

Thomas

et al. 2009

Statistics in Medicine

Self Cite

1,724

1,323

View full text Add to dashboard Cite

show abstract

“…The covariates included linear and quadratic transformations of 11 covariates. We used Bayesian model averaging, implemented via reversible jump Markov chain Monte Carlo (Lunn et al 2009), to calculate posterior probabilities that linear and quadratic terms had nonzero effects [Pr(b 6 ¼ 0)] and to estimate modelaveraged coefficients, which account for uncertainty in model structure (Wintle et al 2003). We fitted two models for each of six species.…”

Section: Methodsmentioning

confidence: 99%

Projecting current and future location, quality, and connectivity of habitat for breeding birds in the Great Basin

et al. 2014

View full text Add to dashboard Cite

. 2014. Projecting current and future location, quality, and connectivity of habitat for breeding birds in the Great Basin. Ecosphere 5(7):82. http://dx.doi. org/10.1890/ES13-00387.1Abstract. We estimated the current location, quality, and connectivity of habitat for 50 species of breeding birds in four mountain ranges in the central Great Basin (Lander, Nye, and Eureka Counties, Nevada) and projected the future location, quality, and connectivity of habitat for these species given different scenarios of climate-induced land-cover change. In the United States, such models are relevant to federally mandated management of wild animals by state-level agencies. We sampled birds during the breeding seasons of 2001-2009 with fixed-radius point counts. For each species, we used boosted regression trees to model incidence (proportion of years a location was surveyed in which the species was present) as a function of topography and current land cover and climate. To assess model fit, we calculated the proportion of binomial deviance explained. We used cross-validation to estimate the predictive accuracy of the models. We applied the conservation planning program Zonation to identify locations where incidences of multiple species were maximized through time given current land cover and two scenarios of land-cover change, expansion of pinyon-juniper woodland into sagebrush shrubsteppe and contraction of riparian woodland. Models based on a set of 13 covariates derived from remotely sensed data had some predictive capacity for 41 of 50 species. Model outputs suggested substantial changes in amount of habitat for many species following projected expansion of pinyon-juniper woodland, but less pronounced changes following projected contraction of riparian woodland. Zonation analyses indicated that the spatial distribution of the highest-quality habitat for the avian assemblage was relatively consistent through time under both scenarios. Breeding birds in the Great Basin commonly are grouped in management plans on the basis of their general association with land-cover classes such as pinyon-juniper woodland, sagebrush shrubsteppe, and riparian woodland. However, even within these groups, the environmental attributes that explained a high proportion of variation in species' incidences and the projected responses to different scenarios of land-cover change varied considerably among species.

show abstract

“…The approach is a Bayesian variant of multivariate adaptive regression splines (Friedman, 1991). Model selection (that is, the identification of which variables to include) was performed using reversible-jump Markov chain Monte Carlo that estimates the posterior probability that a given predictor variable has an association with the response variable, accounting for linear, quadratic and cubic associations (Lunn et al, 2006(Lunn et al, , 2009. River catchment and sampling site were included as clustering variables (given exchangeable priors; equivalent to random effects in a standard mixed model) to control for population history (for example, founder effects and drift) and unmeasured environmental variables by accounting for variation in HL among sites and catchments that was not explained by the remaining predictor variables.…”

Section: Bayesian Modelmentioning

confidence: 99%

Identifying environmental correlates of intraspecific genetic variation

et al. 2016

View full text Add to dashboard Cite

Genetic variation is critical to the persistence of populations and their capacity to adapt to environmental change. The distribution of genetic variation across a species' range can reveal critical information that is not necessarily represented in species occurrence or abundance patterns. We identified environmental factors associated with the amount of intraspecific, individual-based genetic variation across the range of a widespread freshwater fish species, the Murray cod Maccullochella peelii. We used two different approaches to statistically quantify the relative importance of predictor variables, allowing for nonlinear relationships: a random forest model and a Bayesian approach. The latter also accounted for population history. Both approaches identified associations between homozygosity by locus and both disturbance to the natural flow regime and mean annual flow. Homozygosity by locus was negatively associated with disturbance to the natural flow regime, suggesting that river reaches with more disturbed flow regimes may support larger, more genetically diverse populations. Our findings are consistent with the hypothesis that artificially induced perennial flows in regulated channels may provide greater and more consistent habitat and reduce the frequency of population bottlenecks that can occur frequently under the highly variable and unpredictable natural flow regime of the system. Although extensive river regulation across eastern Australia has not had an overall positive effect on Murray cod numbers over the past century, regulation may not represent the primary threat to Murray cod survival. Instead, pressures other than flow regulation may be more critical to the persistence of Murray cod (for example, reduced frequency of large floods, overfishing and chemical pollution).

show abstract

Generic reversible jump MCMC using graphical models

Cited by 86 publications

References 54 publications

The BUGS project: Evolution, critique and future directions

The BUGS project: Evolution, critique and future directions

Projecting current and future location, quality, and connectivity of habitat for breeding birds in the Great Basin

Identifying environmental correlates of intraspecific genetic variation

Contact Info

Product

Resources

About