Dataset Information

GeFaST: An improved method for OTU assignment by generalising Swarm's fastidious clustering approach.

ABSTRACT: BACKGROUND:Massive genomic data sets from high-throughput sequencing allow for new insights into complex biological systems such as microbial communities. Analyses of their diversity and structure are typically preceded by clustering millions of 16S rRNA gene sequences into OTUs. Swarm introduced a new clustering strategy which addresses important conceptual and performance issues of the popular de novo clustering approach. However, some parts of the new strategy, e.g. the fastidious option for increased clustering quality, come with their own restrictions. RESULTS:In this paper, we present the new exact, alignment-based de novo clustering tool GeFaST, which implements a generalisation of Swarm's fastidious clustering. Our tool extends the fastidious option to arbitrary clustering thresholds and allows to adjust its greediness. GeFaST was evaluated on mock-community and natural data and achieved higher clustering quality and performance for small to medium clustering thresholds compared to Swarm and other de novo tools. Clustering with GeFaST was between 6 and 197 times as fast as with Swarm, while the latter required up to 38% less memory for non-fastidious clustering but at least three times as much memory for fastidious clustering. CONCLUSIONS:GeFaST extends the scope of Swarm's clustering strategy by generalising its fastidious option, thereby allowing for gains in clustering quality, and by increasing its performance (especially in the fastidious case). Our evaluations showed that GeFaST has the potential to leverage the use of the (fastidious) clustering strategy for higher thresholds and on larger data sets.

SUBMITTER: Muller R

PROVIDER: S-EPMC6134716 | biostudies-literature | 2018 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

GeFaST: An improved method for OTU assignment by generalising Swarm's fastidious clustering approach.

Müller Robert R Nebel Markus E ME

BMC bioinformatics 20180912 1

<h4>Background</h4>Massive genomic data sets from high-throughput sequencing allow for new insights into complex biological systems such as microbial communities. Analyses of their diversity and structure are typically preceded by clustering millions of 16S rRNA gene sequences into OTUs. Swarm introduced a new clustering strategy which addresses important conceptual and performance issues of the popular de novo clustering approach. However, some parts of the new strategy, e.g. the fastidious opt ...[more]

PMID: 30208838

Similar Datasets

Project description:BackgroundHigh-throughput sequencing has become an essential technology in life science research. Despite continuous improvements in technology, the produced sequences are still not entirely accurate. Consequently, the sequences are usually equipped with error probabilities. The quality information is already employed to find better solutions to a number of bioinformatics problems (e.g. read mapping). Data processing pipelines benefit in particular (especially when incorporating the quality information early), since enhanced outcomes of one step can improve all subsequent ones. Preprocessing steps, thus, quite regularly consider the sequence quality to fix errors or discard low-quality data. Other steps, however, like clustering sequences into operational taxonomic units (OTUs), a common task in the analysis of microbial communities, are typically performed without making use of the available quality information.ResultsIn this paper, we present quality-aware clustering methods inspired by quality-weighted alignments and model-based denoising, and explore their applicability to OTU clustering. We implemented the quality-aware methods in a revised version of our de novo clustering tool GeFaST and evaluated their clustering quality and performance on mock-community data sets. Quality-weighted alignments were able to improve the clustering quality of GeFaST by up to 10%. The examination of the model-supported methods provided a more diverse picture, hinting at a narrower applicability, but they were able to attain similar improvements. Considering the quality information enlarged both runtime and memory consumption, even though the increase of the former depended heavily on the applied method and clustering threshold.ConclusionsThe quality-aware methods expand the iterative, de novo clustering approach by new clustering and cluster refinement methods. Our results indicate that OTU clustering constitutes yet another analysis step benefiting from the integration of quality information. Beyond the shown potential, the quality-aware methods offer a range of opportunities for fine-tuning and further extensions.

Project description:We present a performance-optimized algorithm, subsampled open-reference OTU picking, for assigning marker gene (e.g., 16S rRNA) sequences generated on next-generation sequencing platforms to operational taxonomic units (OTUs) for microbial community analysis. This algorithm provides benefits over de novo OTU picking (clustering can be performed largely in parallel, reducing runtime) and closed-reference OTU picking (all reads are clustered, not only those that match a reference database sequence with high similarity). Because more of our algorithm can be run in parallel relative to "classic" open-reference OTU picking, it makes open-reference OTU picking tractable on massive amplicon sequence data sets (though on smaller data sets, "classic" open-reference OTU clustering is often faster). We illustrate that here by applying it to the first 15,000 samples sequenced for the Earth Microbiome Project (1.3 billion V4 16S rRNA amplicons). To the best of our knowledge, this is the largest OTU picking run ever performed, and we estimate that our new algorithm runs in less than 1/5 the time than would be required of "classic" open reference OTU picking. We show that subsampled open-reference OTU picking yields results that are highly correlated with those generated by "classic" open-reference OTU picking through comparisons on three well-studied datasets. An implementation of this algorithm is provided in the popular QIIME software package, which uses uclust for read clustering. All analyses were performed using QIIME's uclust wrappers, though we provide details (aided by the open-source code in our GitHub repository) that will allow implementation of subsampled open-reference OTU picking independently of QIIME (e.g., in a compiled programming language, where runtimes should be further reduced). Our analyses should generalize to other implementations of these OTU picking algorithms. Finally, we present a comparison of parameter settings in QIIME's OTU picking workflows and make recommendations on settings for these free parameters to optimize runtime without reducing the quality of the results. These optimized parameters can vastly decrease the runtime of uclust-based OTU picking in QIIME.

Dataset Information

GeFaST: An improved method for OTU assignment by generalising Swarm's fastidious clustering approach.

Publications

GeFaST: An improved method for OTU assignment by generalising Swarm's fastidious clustering approach.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets