Dataset Information

NG-meta-profiler: fast processing of metagenomes using NGLess, a domain-specific language.

ABSTRACT: BACKGROUND:Shotgun metagenomes contain a sample of all the genomic material in an environment, allowing for the characterization of a microbial community. In order to understand these communities, bioinformatics methods are crucial. A common first step in processing metagenomes is to compute abundance estimates of different taxonomic or functional groups from the raw sequencing data. Given the breadth of the field, computational solutions need to be flexible and extensible, enabling the combination of different tools into a larger pipeline. RESULTS:We present NGLess and NG-meta-profiler. NGLess is a domain specific language for describing next-generation sequence processing pipelines. It was developed with the goal of enabling user-friendly computational reproducibility. It provides built-in support for many common operations on sequencing data and is extensible with external tools with configuration files. Using this framework, we developed NG-meta-profiler, a fast profiler for metagenomes which performs sequence preprocessing, mapping to bundled databases, filtering of the mapping results, and profiling (taxonomic and functional). It is significantly faster than either MOCAT2 or htseq-count and (as it builds on NGLess) its results are perfectly reproducible. CONCLUSIONS:NG-meta-profiler is a high-performance solution for metagenomics processing built on NGLess. It can be used as-is to execute standard analyses or serve as the starting point for customization in a perfectly reproducible fashion. NGLess and NG-meta-profiler are open source software (under the liberal MIT license) and can be downloaded from https://ngless.embl.de or installed through bioconda.

SUBMITTER: Coelho LP

PROVIDER: S-EPMC6547473 | biostudies-literature | 2019 Jun

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

NG-meta-profiler: fast processing of metagenomes using NGLess, a domain-specific language.

Coelho Luis Pedro LP Alves Renato R Monteiro Paulo P Huerta-Cepas Jaime J Freitas Ana Teresa AT Bork Peer P

Microbiome 20190603 1

<h4>Background</h4>Shotgun metagenomes contain a sample of all the genomic material in an environment, allowing for the characterization of a microbial community. In order to understand these communities, bioinformatics methods are crucial. A common first step in processing metagenomes is to compute abundance estimates of different taxonomic or functional groups from the raw sequencing data. Given the breadth of the field, computational solutions need to be flexible and extensible, enabling the ...[more]

PMID: 31159881

Similar Datasets

Project description:BackgroundThere has been increasing interest in machine learning based natural language processing (NLP) methods in radiology; however, models have often used word embeddings trained on general web corpora due to lack of a radiology-specific corpus.PurposeWe examined the potential of Radiopaedia to serve as a general radiology corpus to produce radiology specific word embeddings that could be used to enhance performance on a NLP task on radiological text.Materials and methodsEmbeddings of dimension 50, 100, 200, and 300 were trained on articles collected from Radiopaedia using a GloVe algorithm and evaluated on analogy completion. A shallow neural network using input from either our trained embeddings or pre-trained Wikipedia 2014 + Gigaword 5 (WG) embeddings was used to label the Radiopaedia articles. Labeling performance was evaluated based on exact match accuracy and Hamming loss. The McNemar's test with continuity and the Benjamini-Hochberg correction and a 5×2 cross validation paired two-tailed t-test were used to assess statistical significance.ResultsFor accuracy in the analogy task, 50-dimensional (50-D) Radiopaedia embeddings outperformed WG embeddings on tumor origin analogies (p < 0.05) and organ adjectives (p < 0.01) whereas WG embeddings tended to outperform on inflammation location and bone vs. muscle analogies (p < 0.01). The two embeddings had comparable performance on other subcategories. In the labeling task, the Radiopaedia-based model outperformed the WG based model at 50, 100, 200, and 300-D for exact match accuracy (p < 0.001, p < 0.001, p < 0.01, and p < 0.05, respectively) and Hamming loss (p < 0.001, p < 0.001, p < 0.01, and p < 0.05, respectively).ConclusionWe have developed a set of word embeddings from Radiopaedia and shown that they can preserve relevant medical semantics and augment performance on a radiology NLP task. Our results suggest that the cultivation of a radiology-specific corpus can benefit radiology NLP models in the future.

Project description:Polymerase chain reaction (PCR) is an in vitro technology in molecular genetics that progressively amplifies minimal copies of short DNA sequences in a fast and inexpensive manner. However, PCR performance is sensitive to suboptimal processing conditions. Compromised PCR conditions lead to artifacts and bias that downgrade the discriminatory power and reproducibility of the results. Promising attempts to resolve the PCR performance optimization issue have been guided by quality improvement tactics adopted in the past for industrial trials. Thus, orthogonal arrays (OAs) have been employed to program quick-and-easy structured experiments. Profiling of influences facilitates the quantification of effects that may counteract the detectability of amplified DNA fragments. Nevertheless, the attractive feature of reducing greatly the amount of work and expenditures by planning trials with saturated-unreplicated OA schemes is known to be relinquished in the subsequent analysis phase. This is because of an inherent incompatibility of ordinary multi-factorial comparison techniques to convert small yet dense datasets. Treating unreplicated-saturated data with either the analysis of variance (ANOVA) or regression models destroys the information extraction process. Both of those mentioned approaches are rendered blind to error since the examined effects absorb all available degrees of freedom. Therefore, in lack of approximating an experimental uncertainty, any outcome interpretation is rendered subjective. We propose a profiling method that permits the non-linear maximization of amplicon resolution by eliminating the necessity for direct error estimation. Our approach is distribution-free, calibration-free, simulation-free and sparsity-free with well-known power properties. It is also user-friendly by promoting rudimentary analytics. Testing our method on published amplicon count data, we found that the preponderant effect is the concentration of MgCl2 (p<0.05) followed by the primer content (p<0.1) whilst the effects due to either the content of the deoxynucleotide (dNTP) or DNA remained dormant (p>0.1). Comparison of the proposed method with other stochastic approaches is also discussed. Our technique is expected to have extensive applications in genetics and biotechnology where there is a demand for cheap, expedient, and robust information.

Dataset Information

NG-meta-profiler: fast processing of metagenomes using NGLess, a domain-specific language.

Publications

NG-meta-profiler: fast processing of metagenomes using NGLess, a domain-specific language.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets