Biology subjects

Kaehler, B. D.

Publications and source records attributed to Kaehler, B. D..

4 recordsLinked to original sources

bioRxiv2022

The Inherited Rate Matrix algorithm for phylogenetic model selection for non-stationary Markov processes

In phylogenetic reconstruction, substitution models that assume nucleotide frequencies do not change through time are in very widespread use. DNA sequences can exhibit marked compositional difference. Such compositional heterogeneity implies sequence divergence has arisen from substitution processes that differ between the lineages. Accordingly, there has been considerable interest in exploring the non-stationary model class, which is capable of generating divergent sequence composition. These models are typically parameter rich and this richness is further compounded by the use of time-heterogeneous variants in order to capture compositional heterogeneity. This illuminates two barriers limiting the widespread application of the non-stationary model class: both the compute-time required to fit a model and the risk of model over-fitting increase with the number of parameters and the extent of time-heterogeneity. In this study, we address these issues with a novel model selection algorithm we term the Inherited Rate Matrix algorithm (hereafter IRM). This approach is based on the notion that a species inherits the substitution tendencies of its ancestor. We further present the non-stationary heterogeneous across lineages model (hereafter ns-HAL algorithm) which extends the HAL algorithm of Jayaswal et al. [2014] to the general nucleotide Markov process. The IRM algorithm substantially reduces the complexity of identifying a sufficient solution to the problem of time-heterogeneous substitution processes across lineages. We also address the issue of reducing the computing time with development of a constrained-optimisation approach for the IRM algorithm (fast-IRM). Our algorithms are implemented in Python 3. From a simulation study based on 2nd codon position genome sequences of yeast, we establish that IRM is significantly more accurate than both ns-HAL and heterogeneity in the substitution process across lineages (hereafter HAL) for close and dispersed sequences. IRM is up to 70x faster than ns-HAL while fast-IRM is around 140x faster. fast-IRM also showed a marked speed improvement over a C++ implementation of HAL. Our comparison of the accuracy of IRM with fast-IRM showed no difference with identical inferences made for all data sets. These two algorithms greatly improve the compute time for model selection of a non-stationary process, increasing the suite of problems to which this important substitution model class can be applied.

bioinformatics↗

bioRxiv2020

RESCRIPt: Reproducible sequence taxonomy reference database management for the masses

BackgroundNucleotide sequence and taxonomy reference databases are critical resources for widespread applications including marker-gene and metagenome sequencing for microbiome analysis, diet metabarcoding, and environmental DNA (eDNA) surveys. Reproducibly generating, managing, using, and evaluating nucleotide sequence and taxonomy reference databases creates a significant bottleneck for researchers aiming to generate custom sequence databases. Furthermore, database composition drastically influences results, and lack of standardizations limits cross-study comparisons. To address these challenges, we developed RESCRIPt, a software package for reproducible generation and management of reference sequence taxonomy databases, including dedicated functions that streamline creating databases from popular sources, and functions for evaluating, comparing, and interactively exploring qualitative and quantitative characteristics across reference databases. ResultsTo highlight the breadth and capabilities of RESCRIPt, we provide several examples for working with popular databases for microbiome profiling (SILVA, Greengenes, NCBI-RefSeq, GTDB), eDNA, and diet metabarcoding surveys (BOLD, GenBank), as well as for genome comparison. We show that bigger is not always better, and reference databases with standardized taxonomies and those that focus on type strains have quantitative advantages, though may not be appropriate for all use cases. Most databases appear to benefit from some curation (quality filtering), though sequence clustering appears detrimental to database quality. Finally, we demonstrate the breadth and extensibility of RESCRIPt for reproducible workflows with a comparison of global hepatitis genomes. ConclusionsRESCRIPt provides tools to democratize the process of reference database acquisition and management, enabling researchers to reproducibly and transparently create reference materials for diverse research applications. RESCRIPt is released under a permissive BSD-3 license at https://github.com/bokulich-lab/RESCRIPt.

bioinformatics↗

bioRxiv2018

Species-level microbial sequence classification is improved by source-environment information

Popular naive Bayes taxonomic classifiers for amplicon sequences assume that all species in the reference database are equally likely to be observed. We demonstrate that classification accuracy degrades linearly with the degree to which that assumption is violated, and in practice it is always violated. By incorporating environment-specific taxonomic abundance information, we demonstrate that species-level resolution is attainable.

bioinformatics↗

bioRxiv2018

q2-sample-classifier: machine-learning tools for microbiome classification and regression

Microbiome studies often aim to predict outcomes or differentiate samples based on their microbial compositions, tasks that can be efficiently performed by supervised learning methods. Here we present a benchmark comparison of supervised learning classifiers and regressors implemented in scikit-learn, a Python-based machine-learning library. We additionally present q2-sample-classifier, a plugin for the QIIME 2 microbiome bioinformatics framework, that facilitates application of the scikit-learn classifiers to microbiome data. Random forest, extra trees, and gradient boosting models demonstrate the highest performance for both supervised classification and regression of microbiome data. Automated feature selection and hyperparameter tuning enhance performance of most methods but may not be necessary under all circumstances. The q2-sample-classifier plugin makes these methods more accessible and interpretable to a broad audience of microbiologists, clinicians, and others who wish to utilize supervised learning methods for predicting sample characteristics based on microbiome composition. The q2-sample-classifier source code is available at https://github.com/qiime2/q2-sample-classifier. It is released under a BSD-3-Clause license, and is freely available including for commercial use.

bioinformatics↗