Recent developments in marginal likelihood estimation for model selection in the field of Bayesian phylogenetics and molecular evolution have emphasized the poor performance of the harmonic mean estimator (HME). Although these studies have shown the merits of new approaches applied to standard normally distributed examples and small real-world data sets, not much is currently known concerning the performance and computational issues of these methods when fitting complex evolutionary and population genetic models to empirical real-world data sets. Further, these approaches have not yet seen widespread application in the field due to the lack of implementations of these computationally demanding techniques in commonly used phylogenetic packages. We here investigate the performance of some of these new marginal likelihood estimators, specifically, path sampling (PS) and stepping-stone (SS) sampling for comparing models of demographic change and relaxed molecular clocks, using synthetic data and real-world examples for which unexpected inferences were made using the HME. Given the drastically increased computational demands of PS and SS sampling, we also investigate a posterior simulation-based analogue of Akaike's information criterion (AIC) through Markov chain Monte Carlo (MCMC), a model comparison approach that shares with the HME the appealing feature of having a low computational overhead over the original MCMC analysis. We confirm that the HME systematically overestimates the marginal likelihood and fails to yield reliable model classification and show that the AICM performs better and may be a useful initial evaluation of model choice but that it is also, to a lesser degree, unreliable. We show that PS and SS sampling substantially outperform these estimators and adjust the conclusions made concerning previous analyses for the three real-world data sets that we reanalyzed. The methods used in this article are now available in BEAST, a powerful user-friendly software package to perform Bayesian evolutionary analyses.

Objective: A history of periodontal disease and the presence of circulating antibodies to selected oral pathogens have been associated with increased risk of pancreatic cancer; however, direct relationships of oral microbes with pancreatic cancer have not been evaluated in prospective studies. We examine the relationship of oral microbiota with subsequent risk of pancreatic cancer in a large nested case–control study. Design: We selected 361 incident adenocarcinoma of pancreas and 371 matched controls from two prospective cohort studies, the American Cancer Society Cancer Prevention Study II and the National Cancer Institute Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial. From pre-diagnostic oral wash samples, we characterised the composition of the oral microbiota using bacterial 16S ribosomal RNA (16S rRNA) gene sequencing. The associations between oral microbiota and risk of pancreatic cancer, controlling for the random effect of cohorts and other covariates, were examined using traditional and L1-penalised least absolute shrinkage and selection operator logistic regression. Results: Carriage of oral pathogens, Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans, were associated with higher risk of pancreatic cancer (adjusted OR for presence vs absence=1.60 and 95% CI 1.15 to 2.22; OR=2.20 and 95% CI 1.16 to 4.18, respectively). Phylum Fusobacteria and its genus Leptotrichia were associated with decreased pancreatic cancer risk (OR per per cent increase of relative abundance=0.94 and 95% CI 0.89 to 0.99; OR=0.87 and 95% CI 0.79 to 0.95, respectively). Risks related to these phylotypes remained after exclusion of cases that developed within 2 years of sample collection, reducing the likelihood of reverse causation in this prospective study. Conclusions: This study provides supportive evidence that oral microbiota may play a role in the aetiology of pancreatic cancer.

Oral microbiome dysbiosis is associated with oral disease and potentially with systemic diseases; however, the determinants of these microbial imbalances are largely unknown. In a study of 1204 US adults, we assessed the relationship of cigarette smoking with the oral microbiome. 16S rRNA gene sequencing was performed on DNA from oral wash samples, sequences were clustered into operational taxonomic units (OTUs) using QIIME and metagenomic content was inferred using PICRUSt. Overall oral microbiome composition differed between current and non-current (former and never) smokers (P<0.001). Current smokers had lower relative abundance of the phylum Proteobacteria (4.6%) compared with never smokers (11.7%) (false discovery rate q=5.2 × 10−7), with no difference between former and never smokers; the depletion of Proteobacteria in current smokers was also observed at class, genus and OTU levels. Taxa not belonging to Proteobacteria were also associated with smoking: the genera Capnocytophaga, Peptostreptococcus and Leptotrichia were depleted, while Atopobium and Streptococcus were enriched, in current compared with never smokers. Functional analysis from inferred metagenomes showed that bacterial genera depleted by smoking were related to carbohydrate and energy metabolism, and to xenobiotic metabolism. Our findings demonstrate that smoking alters the oral microbiome, potentially leading to shifts in functional pathways with implications for smoking-related diseases.

A Family of Languages: English is part of the large Indo-European language family, which includes Celtic, Germanic, Italic, Balto-Slavic, and Indo-Iranian languages. The origin of this family is hotly debated: one hypothesis places the origin north of the Caspian Sea in the Pontic steppes, from where it was disseminated by Kurgan semi-nomadic pastoralists; a second suggests that Anatolia, in modern-day Turkey, is the source, and the language radiated with the spread of agriculture. Bouckaert et al. (p. 957 ) used phylogenetic methods and modeling to assess the geographical spread of the Indo-European language group. The findings support the suggestion that the origin of the language family was indeed Anatolia 7 to 10 thousand years ago—contemporaneous with the spread of agriculture.