Search

MOTIVATION: Advancements in high-throughput genomic sequencing are delivering genomic pathogen data at an unprecedented rate, positioning statistical phylogenetics as a critical tool to monitor infectious diseases globally. This rapid growth spurs the need for efficient inference techniques, such as Hamiltonian Monte Carlo (HMC) in a Bayesian framework, to estimate parameters of these phylogenetic models where the dimensions of the parameters increase with the number of sequences N. HMC requires repeated calculation of the gradient of the data log-likelihood with respect to (wrt) all branch-length-specific (BLS) parameters that traditionally takes O(N2) operations using the standard pruning algorithm. A recent study proposes an approach to calculate this gradient in O(N), enabling researchers to take advantage of gradient-based samplers such as HMC. The CPU implementation of this approach makes the calculation of the gradient computationally tractable for nucleotide-based models but falls short in performance for larger state-space size models, such as Markov-modulated and codon models. Here, we describe novel massively parallel algorithms to calculate the gradient of the log-likelihood wrt all BLS parameters that take advantage of graphics processing units (GPUs) and result in many fold higher speedups over previous CPU implementations. RESULTS: We benchmark these GPU algorithms on three computing systems using three evolutionary inference examples exploring complete genomes from 997 dengue viruses, 62 carnivore mitochondria and 49 yeasts, and observe a >128-fold speedup over the CPU implementation for codon-based models and >8-fold speedup for nucleotide-based models. As a practical demonstration, we also estimate the timing of the first introduction of West Nile virus into the continental Unites States under a codon model with a relaxed molecular clock from 104 full viral genomes, an inference task previously intractable. AVAILABILITY AND IMPLEMENTATION: We provide an implementation of our GPU algorithms in BEAGLE v4.0.0 (https://github.com/beagle-dev/beagle-lib), an open-source library for statistical phylogenetics that enables parallel calculations on multi-core CPUs and GPUs. We employ a BEAGLE-implementation using the Bayesian phylogenetics framework BEAST (https://github.com/beast-dev/beast-mcmc).

SummaryMutations sometimes increase contagiousness for evolving pathogens. During an epidemic, scientists use viral genome data to infer a shared evolutionary history and connect this history to geographic spread. We propose a model that directly relates a pathogen's evolution to its spatial contagion dynamics-effectively combining the two epidemiological paradigms of phylogenetic inference and self-exciting process modeling-and apply this phylogenetic Hawkes process to a Bayesian analysis of 23 421 viral cases from the 2014 to 2016 Ebola outbreak in West Africa. The proposed model is able to detect individual viruses with significantly elevated rates of spatiotemporal propagation for a subset of 1610 samples that provide genome data. Finally, to facilitate model application in big data settings, we develop massively parallel implementations for the gradient and Hessian of the log-likelihood and apply our high-performance computing framework within an adaptively pre-conditioned Hamiltonian Monte Carlo routine.Supplementary informationSupplementary data are available at Bioinformatics online.

MotivationThe potentially low precision associated with the geographic origin of sampled sequences represents an important limitation for spatially explicit (i.e. continuous) phylogeographic inference of fast-evolving pathogens such as RNA viruses. A substantial proportion of publicly available sequences is geo-referenced at broad spatial scale such as the administrative unit of origin, rather than more precise locations (e.g. geographic coordinates). Most frequently, such sequences are either discarded prior to continuous phylogeographic inference or arbitrarily assigned to the geographic coordinates of the centroid of their administrative area of origin for lack of a better alternative.ResultsWe here implement and describe a new approach that allows to incorporate heterogeneous prior sampling probabilities over a geographic area. External data, such as outbreak locations, are used to specify these prior sampling probabilities over a collection of sub-polygons. We apply this new method to the analysis of highly pathogenic avian influenza H5N1 clade data in the Mekong region. Our method allows to properly include, in continuous phylogeographic analyses, H5N1 sequences that are only associated with large administrative areas of origin and assign them with more accurate locations. Finally, we use continuous phylogeographic reconstructions to analyse the dispersal dynamics of different H5N1 clades and investigate the impact of environmental factors on lineage dispersal velocities.Availability and implementationOur new method allowing heterogeneous sampling priors for continuous phylogeographic inference is implemented in the open-source multi-platform software package BEAST 1.10.Supplementary informationSupplementary data are available at Bioinformatics online.

MotivationAdvances in sequencing technology continue to deliver increasingly large molecular sequence datasets that are often heavily partitioned in order to accurately model the underlying evolutionary processes. In phylogenetic analyses, partitioning strategies involve estimating conditionally independent models of molecular evolution for different genes and different positions within those genes, requiring a large number of evolutionary parameters that have to be estimated, leading to an increased computational burden for such analyses. The past two decades have also seen the rise of multi-core processors, both in the central processing unit (CPU) and Graphics processing unit processor markets, enabling massively parallel computations that are not yet fully exploited by many software packages for multipartite analyses.ResultsWe here propose a Markov chain Monte Carlo (MCMC) approach using an adaptive multivariate transition kernel to estimate in parallel a large number of parameters, split across partitioned data, by exploiting multi-core processing. Across several real-world examples, we demonstrate that our approach enables the estimation of these multipartite parameters more efficiently than standard approaches that typically use a mixture of univariate transition kernels. In one case, when estimating the relative rate parameter of the non-coding partition in a heterochronous dataset, MCMC integration efficiency improves by > 14-fold.Availability and implementationOur implementation is part of the BEAST code base, a widely used open source software package to perform Bayesian phylogenetic inference.Contactguy.baele@kuleuven.be.Supplementary informationSupplementary data are available at Bioinformatics online.

SummarySPREAD is a user-friendly, cross-platform application to analyze and visualize Bayesian phylogeographic reconstructions incorporating spatial-temporal diffusion. The software maps phylogenies annotated with both discrete and continuous spatial information and can export high-dimensional posterior summaries to keyhole markup language (KML) for animation of the spatial diffusion through time in virtual globe software. In addition, SPREAD implements Bayes factor calculation to evaluate the support for hypotheses of historical diffusion among pairs of discrete locations based on Bayesian stochastic search variable selection estimates. SPREAD takes advantage of multicore architectures to process large joint posterior distributions of phylogenies and their spatial diffusion and produces visualizations as compelling and interpretable statistical summaries for the different spatial projections.AvailabilitySPREAD is licensed under the GNU Lesser GPL and its source code is freely available as a GitHub repository: https://github.com/phylogeography/SPREAD CONTACT: filip.bielejec@rega.kuleuven.be.

Motivation: Statistical analyses of phylogenetic data culminate in uncertain estimates of underlying model parameters. Lack of additional data hinders the ability to reduce this uncertainty, as the original phylogenetic dataset is often complete, containing the entire gene or genome information available for the given set of taxa. Informative priors in a Bayesian analysis can reduce posterior uncertainty; however, publicly available phylogenetic software specifies vague priors for model parameters by default. We build objective and informative priors using hierarchical random effect models that combine additional datasets whose parameters are not of direct interest but are similar to the analysis of interest. [ABSTRACT FROM PUBLISHER]