Inference Software Research Articles

Teaching Introductory Statistics with DataCamp

We designed a sequence of courses for the DataCamp online learning platform that approximates the content of a typical introductory statistics course. We discuss the design and implementation of these courses and illustrate how they can be successfully integrated into a brick-and-mortar class. We reflect on the process of creating content for online consumers, ruminate on the pedagogical considerations we faced, and describe an R package for statistical inference that became a by-product of this development process. We discuss the pros and cons of creating the course sequence and express our view that some aspects were particularly problematic. The issues raised should be relevant to nearly all statistics instructors. Supplementary materials for this article are available online.

Prevalence and Identification of Major Ticks in Cattle in Damot Gale Woreda, Wolaita Zone, SNNPR, Ethiopia

Across-sectional study was designed and conducted in Damot Gale district of Wolaita zone, south Ethiopia. The major objectives of the research were: to determine the prevalence of major ixodid ticks; to identify the collected ticks into their respective species; and to assess the risk factor influencing the prevalence of tick infestation in the study area. For this purpose, 384 heads of cattle maintained under traditional management system were selected using simple random sampling statistical technique. All visible ticks were collected from half body part of selected cattle, preserved in70% alcohol and transported to WSU-SVM for subsequent identification into species, based on key standards. Questionnaire survey was conducted to collect baseline information regarding farmers’ perceptions of the effects of ticks on livestock production. The collected data were analyzed using SPSS software (Version20.1) for descriptive and inference purposed. A total of 264 ixodid ticks were collected, which belonged to of Boophilus (59.85%), Rhippicephalus (25.38%) and Amblyomma (14.77%) genera, with the overall prevalence of tick infestation to be about 6%. At species level, Boophilus decoloratus (59.85), Rhippicephalusevertsievertsi (25.38%), Amblyomma variegatum (10.61%) and Amblyommagemma (4.16%) were the hard tick species identified during the survey period. Regarding spatial distribution, these ticks had the highest prevalence at Gacheno and Chocha and the lowest prevalence at Fate district. Regression analysis indicated that the prevalence of tick infestation did not reveal significant statistical difference between male and female sex groups (p>0.05). Nevertheless, adult cattle over the age of three years were more significantly challenged by tick infestation than young cattle (P<0.05). Similarly, the prevalence of tick infestation was significantly higher (p<0.01) in cattle with poor body condition than those with medium and good body condition score. In conclusion, the present research disclosed that tick infestation due to hard tick s is responsible for considerable pathogenic impacts and economic losses due to their deleterious effects on leather industry. Based on the present findings, it is recommended that integrate tick control efforts should be implemented, and the efficacy acaricides should be detected to control the risk of drug resistance in Damot Gale district.

Abstracting container technologies and transfer mechanisms in the Scalable CyberInfrastructure for Artificial Intelligence and Likelihood Free Inference (SCAILFIN) project

High Performance Computing (HPC) facilities provide vast computational power and storage, but generally work on fixed environments designed to address the most common software needs locally, making it challenging for users to bring their own software. To overcome this issue, most HPC facilities have added support for HPC friendly container technologies such as Shifter, Singularity, or Charliecloud. These different container technologies are all compatible with the more popular Docker containers, however the implementation and use of said containers is different for each HPC friendly container technology. These usage differences can make it difficult for an end user to easily submit and utilize different HPC sites without making adjustments to their workflows and software. This issue is exacerbated when attempting to utilize workflow management software between different sites with differing container technologies.The SCAILFIN project aims to develop and deploy artificial intelligence (AI) and likelihood-free inference (LFI) techniques and software using scalable cyberinfrastructure (CI) that span multiple sites. The project has extended the CERN-based REANA framework, a platform designed to enable analysis reusability, and reproducibility while supporting different workflow engine languages, in order to support submission to different HPC facilities. The work presented here focuses on the development of an abstraction layer that allows the support of different container technologies and different transfer protocols for files and directories between the HPC facility and the REANA cluster edge service from the user’s workflow application.

Large-scale HPC deployment of Scalable CyberInfrastructure for Artificial Intelligence and Likelihood Free Inference (SCAILFIN)

The NSF-funded Scalable CyberInfrastructure for Artificial Intelligence and Likelihood Free Inference (SCAILFIN) project aims to develop and deploy artificial intelligence (AI) and likelihood-free inference (LFI) techniques and software using scalable cyberinfrastructure (CI) built on top of existing CI elements. Specifically, the project has extended the CERN-based REANA framework, a cloud-based data analysis platform deployed on top of Kubernetes clusters that was originally designed to enable analysis reusability and reproducibility. REANA is capable of orchestrating extremely complicated multi-step workflows, and uses Kubernetes clusters both for scheduling and distributing container-based workloads across a cluster of available machines, as well as instantiating and monitoring the concrete workloads themselves. This work describes the challenges and development efforts involved in extending REANA and the components that were developed in order to enable large scale deployment on High Performance Computing (HPC) resources. Using the Virtual Clusters for Community Computation (VC3) infrastructure as a starting point, we implemented REANA to work with a number of differing workload managers, including both high performance and high throughput, while simultaneously removing REANA’s dependence on Kubernetes support at the workers level.

Simode: R Package for Statistical Inference of Ordinary Differential Equations using Separable Integral-Matching

In this paper we describe simode: Separable Integral Matching for Ordinary Differential Equations. The statistical methodologies applied in the package focus on several minimization procedures of an integral-matching criterion function, taking advantage of the mathematical structure of the differential equations like separability of parameters from equations. Application of integral based methods to parameter estimation of ordinary differential equations was shown to yield more accurate and stable results comparing to derivative based ones. Linear features such as separability were shown to ease optimization and inference. We demonstrate the functionalities of the package using various systems of ordinary differential equations.

MPF–BML: a standalone GUI-based package for maximum entropy model inference

Learning underlying correlation patterns in data is a central problem across scientific fields. Maximum entropy models present an important class of statistical approaches for addressing this problem. However, accurately and efficiently inferring model parameters are a major challenge, particularly for modern high-dimensional applications such as in biology, for which the number of parameters is enormous. Previously, we developed a statistical method, minimum probability flow-Boltzmann Machine Learning (MPF-BML), for performing fast and accurate inference of maximum entropy model parameters, which was applied to genetic sequence data to estimate the fitness landscape for the surface proteins of human immunodeficiency virus and hepatitis C virus. To facilitate seamless use of MPF-BML and encourage more widespread application to data in diverse fields, we present a standalone cross-platform package of MPF-BML which features an easy-to-use graphical user interface. The package only requires the input data (protein sequence data or data of multiple configurations of a complex system with large number of variables) and returns the maximum entropy model parameters. The MPF-BML software is publicly available under the MIT License at https://github.com/ahmedaq/MPF-BML-GUI. Supplementary data are available at Bioinformatics online.

PYLFIRE: Python implementation of likelihood-free inference by ratio estimation.

Likelihood-free inference for simulator-based models is an emerging methodological branch of statistics which has attracted considerable attention in applications across diverse fields such as population genetics, astronomy and economics. Recently, the power of statistical classifiers has been harnessed in likelihood-free inference to obtain either point estimates or even posterior distributions of model parameters. Here we introduce PYLFIRE, an open-source Python implementation of the inference method LFIRE (likelihood-free inference by ratio estimation) that uses penalised logistic regression.PYLFIRE is made available as part of the general ELFI inference software http://elfi.ai to benefit both the user and developer communities for likelihood-free inference.

MCMC4Extremes: an R package for Bayesian inference for extremes and its extensions

The R package, MCMC4Extremes, provides functions which estimate posterior points for extreme-value distributions such as the Generalized Pareto Distribution (GPD), the Generalize Extreme Value (GEV), and the new extension dual Gamma Generalized Extreme Value (GGEV). Inference is performed under the Bayesian paradigm. The Markov chain Monte Carlo (MCMC) method is used to obtain the posterior points. Moreover, the package allows the identification of predictive distributions for the estimated parameters. It performs estimation for return levels plots, and fit measures such as AIC, BIC, and DIC.

Per-sample immunoglobulin germline inference from B cell receptor deep sequencing data.

The collection of immunoglobulin genes in an individual’s germline, which gives rise to B cell receptors via recombination, is known to vary significantly across individuals. In humans, for example, each individual has only a fraction of the several hundred known V alleles. Furthermore, the currently-accepted set of known V alleles is both incomplete (particularly for non-European samples), and contains a significant number of spurious alleles. The resulting uncertainty as to which immunoglobulin alleles are present in any given sample results in inaccurate B cell receptor sequence annotations, and in particular inaccurate inferred naive ancestors. In this paper we first show that the currently widespread practice of aligning each sequence to its closest match in the full set of IMGT alleles results in a very large number of spurious alleles that are not in the sample’s true set of germline V alleles. We then describe a new method for inferring each individual’s germline gene set from deep sequencing data, and show that it improves upon existing methods by making a detailed comparison on a variety of simulated and real data samples. This new method has been integrated into the partis annotation and clonal family inference package, available at https://github.com/psathyrella/partis, and is run by default without affecting overall run time.

OSCR: a spatial capture–recapture R package for inference about spatial ecological processes

Spatial capture–recapture (SCR) methods have become widely applied in ecology. The immediate adoption of SCR is due to the fact that it resolves some major criticisms of traditional capture–recapture methods related to heterogeneity in detectabililty, and the emergence of new technologies (e.g. camera traps, non‐invasive genetics) that have vastly improved our ability to collection spatially explicit observation data on individuals. However, the utility of SCR methods reaches far beyond simply convenience and data availability. SCR presents a formal statistical framework that can be used to test explicit hypotheses about core elements of population and landscape ecology, and has profound implications for how we study animal populations. In this software note, we describe the technical basis and analytical workflow of oSCR, an R package for analyzing spatial encounter history data using a multi‐session sex‐structured likelihood. The impetus for developing oSCR was to create an accessible and transparent analysis tool that allows users to conveniently and intuitively formulate statistical models that map directly to fundamental processes of interest in spatial population ecology (e.g. space use, resource selection, density and connectivity). We have placed an emphasis on creating a transparent and accessible code base that is coupled with a logical workflow that we hope stimulates active participation in further technical developments.

Computational Inference Software for Tetrad Assembly from Randomly Arrayed Yeast Colonies.

We describe an information-theory-based method and associated software for computationally identifying sister spores derived from the same meiotic tetrad. The method exploits specific DNA sequence features of tetrads that result from meiotic centromere and allele segregation patterns. Because the method uses only the genomic sequence, it alleviates the need for tetrad-specific barcodes or other genetic modifications to the strains. Using this method, strains derived from randomly arrayed spores can be efficiently grouped back into tetrads.

Fractional coalescent

An approach to the coalescent, the fractional coalescent (f-coalescent), is introduced. The derivation is based on the discrete-time Cannings population model in which the variance of the number of offspring depends on the parameter α. This additional parameter α affects the variability of the patterns of the waiting times; values of [Formula: see text] lead to an increase of short time intervals, but occasionally allow for very long time intervals. When [Formula: see text], the f-coalescent and the Kingman's n-coalescent are equivalent. The distribution of the time to the most recent common ancestor and the probability that n genes descend from m ancestral genes in a time interval of length T for the f-coalescent are derived. The f-coalescent has been implemented in the population genetic model inference software Migrate Simulation studies suggest that it is possible to accurately estimate α values from data that were generated with known α values and that the f-coalescent can detect potential environmental heterogeneity within a population. Bayes factor comparisons of simulated data with [Formula: see text] and real data (H1N1 influenza and malaria parasites) showed an improved model fit of the f-coalescent over the n-coalescent. The development of the f-coalescent and its inclusion into the inference program Migrate facilitates testing for deviations from the n-coalescent.

Evidence for diffuse molecular gas and dust in the hearts of gamma-ray burst host galaxies

Context. Damped Lyman-α (DLA) absorption-line systems at the redshifts of gamma-ray burst (GRB) afterglows offer a unique way to probe the physical conditions within star-forming galaxies in the early Universe. Aims. Here we built up a large sample of 22 GRBs at redshifts z &gt; 2 observed with VLT/X-shooter in order to determine the abundances of hydrogen, metals, dust, and molecular species. This allows us to study the metallicity and dust depletion effects in the neutral interstellar medium at high redshift and to answer the question of whether (and why) there might be a lack of H2 in GRB-DLAs. Methods. We developed new methods based on the Bayesian inference package, PyMC, to FIT absorption lines and measure the column densities of different metal species as well as atomic and molecular hydrogen. The derived relative abundances are used to FIT dust depletion sequences and determine the dust-to-metals ratio and the host-galaxy intrinsic visual extinction. Additionally, we searched for the absorption signatures of vibrationally-excited H2 and carbon monoxide. Results. We find that there is no lack of H2-bearing GRB-DLAs. We detect absorption lines from molecular hydrogen in 6 out of 22 GRB afterglow spectra, with molecular fractions ranging between f ≃ 5 × 10−5 and f ≃ 0.04, and claim tentative detections in three additional cases. For the remainder of the sample, we measure, depending on S/N, spectral coverage and instrumental resolution, more or less stringent upper limits. The GRB-DLAs in our sample have on average low metallicities, [X/H]¯ ≈ −1.3, comparable to the population of extremely-strong QSO-DLAs (log N(H I) &gt; 21.5). Furthermore, H2-bearing GRB-DLAs are found to be associated with significant dust extinction, AV &gt; 0.1 mag, and dust-to-metals ratios DTM &gt; 0.4, confirming the importance of dust grains for the production of molecules. All these systems exhibit neutral hydrogen column densities log N(H I) &gt; 21.7. The overall fraction of H2 detections in GRB-DLAs is ≥ 27% (41% including tentative detections), which is three to four times larger than in the general QSO-DLA population. For 2 &lt; z &lt; 4, and considering column densities log N(H I) &gt; 21.7, the H2 detection fraction is 60–80% in GRB-DLAs and in extremely strong QSO-DLAs. This is likely due to the fact that both GRB- and QSO-DLAs with high neutral hydrogen column densities are probed by sight-lines with small impact parameters, indicating that the absorbing gas is associated with the inner regions of the absorbing galaxy, where the gas pressure is higher and the conversion of H I to H2 takes place. In the case of GRB hosts, this diffuse molecular gas is located at distances ≳ 500 pc from the GRB and hence is unrelated to the star-forming region where the event occurred.

Cappuccino: Efficient CNN Inference Software Synthesis for Mobile System-on-Chips

Convolutional neural networks (CNNs) exhibit remarkable performance in various machine learning tasks. As sensor-equipped Internet of Things devices permeate into every aspect of modern life, the ability to execute CNN inference, a computationally intensive application, on resource constrained devices has become increasingly important. In this context, we present Cappuccino, a framework for synthesis of efficient inference software targeting mobile system-on-chips (SoCs). We propose techniques for efficient parallelization of CNN inference targeting mobile SoCs, and explore the underlying tradeoffs. Experiments with different CNNs on three mobile devices demonstrate the effectiveness of our approach.

Fast Estimation of Recombination Rates Using Topological Data Analysis.

Accurate estimation of recombination rates is critical for studying the origins and maintenance of genetic diversity. Because the inference of recombination rates under a full evolutionary model is computationally expensive, we developed an alternative approach using topological data analysis (TDA) on genome sequences. We find that this method can analyze datasets larger than what can be handled by any existing recombination inference software, and has accuracy comparable to commonly used model-based methods with significantly less processing time. Previous TDA methods used information contained solely in the first Betti number ([Formula: see text]) of a set of genomes, which aims to capture the number of loops that can be detected within a genealogy. These explorations have proven difficult to connect to the theory of the underlying biological process of recombination, and, consequently, have unpredictable behavior under perturbations of the data. We introduce a new topological feature, which we call ψ, with a natural connection to coalescent models, and present novel arguments relating [Formula: see text] to population genetic models. Using simulations, we show that ψ and [Formula: see text] are differentially affected by missing data, and package our approach as TREE (Topological Recombination Estimator). TREE's efficiency and accuracy make it well suited as a first-pass estimator of recombination rate heterogeneity or hotspots throughout the genome. Our work empirically and theoretically justifies the use of topological statistics as summaries of genome sequences and describes a new, unintuitive relationship between topological features of the distribution of sequence data and the footprint of recombination on genomes.

IDTxl: The Information Dynamics Toolkit xl: a Python package for the efficient analysis of multivariate information dynamics in networks

The Information Dynamics Toolkit xl (IDTxl) is a comprehensive software package for efficient inference of networks and their node dynamics from multivariate time series data using information theory. IDTxl provides functionality to estimate the following measures: 1) For network inference: multivariate transfer entropy (TE)/Granger causality (GC), multivariate mutual information (MI), bivariate TE/GC, bivariate MI 2) For analysis of node dynamics: active information storage (AIS), partial information decomposition (PID) IDTxl implements estimators for discrete and continuous data with parallel computing engines for both GPU and CPU platforms. Written for Python3.4.3+.

MontePython 3: Boosted MCMC sampler and other features

Abstract MontePython is a parameter inference package for cosmology. We present the latest development of the code over the past couple of years. We explain, in particular, two new ingredients both contributing to improve the performance of Metropolis–Hastings sampling: an adaptation algorithm for the jumping factor, and a calculation of the inverse Fisher matrix, which can be used as a proposal density. We present several examples to show that these features speed up convergence and can save many hundreds of CPU-hours in the case of difficult runs, with a poor prior knowledge of the covariance matrix. We also summarize all the functionalities of MontePython in the current release, including new likelihoods and plotting options.

Phylogenetic Studies.

Phylogenetics is an important component of the systems biology approach. Knowledge about evolution of the genus Leishmania is essential to understand various aspects of basic biology of these parasites, such as parasite-host or parasite-vector relationships, biogeography, or epidemiology. Here, we present a comprehensive guideline for performing phylogenetic studies based on DNA sequence data, but with principles that can be adapted to protein sequences or other molecular markers. It is presented as a compilation of the most commonly used genetic targets for phylogenetic studies of Leishmania, including their respective primers for amplification and references, as well as details of PCR assays. Guidelines are, then, presented to choose the best targets in relation to the types of samples under study. Finally, and importantly, instructions are given to obtain optimal sequences, alignments, and datasets for the subsequent data analysis and phylogenetic inference. Different bioinformatics methods and software for phylogenetic inference are presented and explained. This chapter aims to provide a compilation of methods and generic guidelines to conduct phylogenetics of Leishmania for nonspecialists.

FactorialHMM: fast and exact inference in factorial hidden Markov models.

Hidden Markov models (HMMs) are powerful tools for modeling processes along the genome. In a standard genomic HMM, observations are drawn, at each genomic position, from a distribution whose parameters depend on a hidden state, and the hidden states evolve along the genome as a Markov chain. Often, the hidden state is the Cartesian product of multiple processes, each evolving independently along the genome. Inference in these so-called Factorial HMMs has a naïve running time that scales as the square of the number of possible states, which by itself increases exponentially with the number of sub-chains; such a running time scaling is impractical for many applications. While faster algorithms exist, there is no available implementation suitable for developing bioinformatics applications. We developed FactorialHMM, a Python package for fast exact inference in Factorial HMMs. Our package allows simulating either directly from the model or from the posterior distribution of states given the observations. Additionally, we allow the inference of all key quantities related to HMMs: (i) the (Viterbi) sequence of states with the highest posterior probability; (ii) the likelihood of the data and (iii) the posterior probability (given all observations) of the marginal and pairwise state probabilities. The running time and space requirement of all procedures is linearithmic in the number of possible states. Our package is highly modular, providing the user with maximal flexibility for developing downstream applications. https://github.com/regevs/factorial_hmm. Supplementary data are available at Bioinformatics online.

Critical steps for computational inference of the 3′-end of novel alleles of immunoglobulin heavy chain variable genes - illustrated by an allele of IGHV3-7

Sequencing of immunoglobulin germline gene loci is a challenging process, e.g. due to their repetitiveness and complexity, hence limiting the insight in the germline gene repertoire of humans and other species. Through next generation sequencing technology, it is possible to generate immunoglobulin transcript data sets large enough to computationally infer the germline genes from which the transcripts originate. Multiple tools for such inference have been developed and they can be used for construction of individual germline gene databases, and for discovery of new immunoglobulin germline genes and alleles. However, there are challenges associated with these methods, many of them related to the biological process through which immunoglobulin coding genes are generated. The junctional diversity introduced during rearrangement of the immunoglobulin heavy chain variable (IGHV), diversity and joining genes specifically complicates the inference of the junction regions, with implications for inference of the 3′-end of IGHV genes. With the aim of coping with such diversity, an inference software package may not be able to identify novel alleles harbouring a difference in these regions compared to their closest relatives in the starting database. In this study, we were able to computationally infer one such previously uncharacterized allele, IGHV3-7*02 A318G. However, this was possible only if a strategy was used in which different variants of IGHV3-7*02 were included in the inference-initiating database. Importantly, the presence of the novel allele, but not the standard IGHV3-7*02 sequence, in the genotype was strongly supported by the actual sequences that were assigned to the allele. We thus showed that the starting database used will impact the germline gene inference process, and that difference in the 3′-end of IGHV genes may remain undetected unless specific, non-standard procedures are used to address this matter. We suggest that inferred genes/alleles should be confirmed e.g. by examination of the nucleotide composition of the 3′-bases of the inference-supporting sequence reads.