Computing Bayes factors using thermodynamic integration.

By Nicolas Lartillot, Hervé Philippe. Systematic Biology, 2006

677 citations recorded by the source corpus.

Subfields: Genetics, Molecular Biology, Paleontology

View Paper: https://www.semanticscholar.org/paper/5bc789277de15c961e7d694a808100193f15f327

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

In Molecular Biology

In Paleontology

Citations in the corpus

Structural Insight into How Bacteria Prevent Interference between Multiple Divergent Type IV Secretion Systems (2015, methodology): “We also analyzed the masked data set with the CAT substitution model, a nonparametric method for modeling site-specific features of sequence evolution (68, 69).”
Improving power posterior estimation of statistical evidence (2012, background): “Lartillot and Philippe (2006) discuss very similar ideas in the phylogenetics literature, although using Simpson’s rule for the numerical integration; they use equally spaced inverse temperatures between 0 and 1.”
Accurate model selection of relaxed molecular clocks in bayesian phylogenetics. (2012, methodology): “We calculate marginal likelihoods using PS (Gelman and Meng 1998; Ogata 1989; Lartillot and Philippe 2006), SS (Xie et al. 2011), HME (Newton and Raftery 1994), and sHME (Redelings and Suchard 2005).”
MrBayes 3.2: Efficient Bayesian Phylogenetic Inference and Model Choice Across a Large Model Space (2012, neutral): “However, there are now considerably more accurate, albeit computationally more demanding, methods (Lartillot and Philippe 2006).”
Bayesian phylogenetic and phylodynamic data integration using BEAST 1.10 (2018, methodology): “BEAST 1.10 now features marginal likelihood estimation (Baele et al. 2012), using path sampling (Gelman and Meng 1998; Lartillot and Philippe 2006) and stepping-stone sampling (Xie et al. 2011), as well as the recently developed generalized stepping-stone sampling (Fan et al. 2011; Baele et al.…”
A taxonomic revision of the Neotropical poison frog genus Ranitomeya (Amphibia: Dendrobatidae) (2011, background): “…the use of harmonic means to calculate the marginal likelihood for model estimation; in particular, harmonic means can favor more parameter-rich models (Lartillot & Philippe 2006) and they can possess a systematic bias capable of significantly skewing Bayes factors (Calderhead & Girolami 2009).”
An adaptive scheduling scheme for calculating Bayes factors with thermodynamic integration using Simpson’s rule (2015, methodology): “The power posterior method exploits ideas from path sampling (Gelman and Meng 1998) in the form of thermodynamic integration and has been introduced in Lartillot and Philippe (2006), Friel and Pettitt (2008).”
Species Tree Inference with BPP Using Genomic Sequences and the Multispecies Coalescent (2018, methodology): “2011), and for calculating the marginal likelihood (Bayes Factors) through thermodynamic integration and Gaussian quadrature (Lartillot and Philippe 2006; Rannala and Yang 2017).”
Evaluating marginal likelihood with thermodynamic integration method and comparison with several other numerical methods (2016, methodology): “This study is focused on two MC approaches that do not use harmonic mean, and they are the thermodynamic integration method [Gelman and Meng, 1998; Neal, 2000; Lartillot and Philippe, 2006; Friel and Pettitt, 2008] and the nested sampling method [Skilling, 2004, 2006].”
Suppression of long-branch attraction artefacts in the animal phylogeny using a site-heterogeneous model (2007, positive): “These results confirm our previous observations based on Bayes factor evaluations [31,41], showing that, in most instances, CAT offers a better statistical fit than WAG.”
Improving marginal likelihood estimation for Bayesian phylogenetic model selection. (2011, methodology): “The method outlined by Lartillot and Philippe (2006) and Friel and Pettitt (2008) differs from the path sampling method defined by Gelman and Meng (1998).”
Phylogeography takes a relaxed random walk in continuous space and time. (2010, methodology): “…in a phylogenetic setting (Suchard et al. 2001; Edwards, Liu et al. 2007; Drummond and Suchard 2008) In borderline cases, more powerful BF estimators are readily available, for example, path sampling (Lartillot and Philippe 2006), and they are currently being implemented in the BEAST framework.”
Empirical profile mixture models for phylogenetic reconstruction (2008, methodology): “The resulting model was shown to provide a better fit than empirical matrices in several cases (Lartillot and Philippe, 2006).”
First Sagittarius A* Event Horizon Telescope Results. IV. Variability, Morphology, and Black Hole Mass (2022, methodology): “To compute the Bayesian evidence, THEMIS uses thermodynamic integration (e.g., Lartillot & Philippe 2006), which computes the log-evidence through ò b= á ñb ( ) ln d ln , 470 1 where á ñbln is the expectation of the log-likelihood taken over the distribution ⎜ ⎟ ⎛ ⎝ ⎞ ⎠p p=b b ( ) ( ) ( ) ( ) ( )(…”