关键词: algorithm amplicon bioinformatics microbiome profiling software

Mesh : Microbiota / genetics RNA, Ribosomal, 16S / genetics Algorithms Bacteria / genetics classification isolation & purification Software Humans High-Throughput Nucleotide Sequencing / methods Phylogeny Computational Biology / methods Sequence Analysis, DNA / methods Molecular Sequence Annotation

来  源:   DOI:10.1128/spectrum.00695-24   PDF(Pubmed)

Abstract:
Amplicon sequencing stands as a cornerstone in microbiome profiling, yet concerns persist regarding its resolution and accuracy. The enhancement of reference databases and annotations marks a new era for 16S rRNA-based profiling. Capitalizing on this potential, we introduce PM-profiler, a novel tool for profiling amplicon short reads. PM-profiler is implemented by C++-based advanced algorithms, such as pre-allocated hash for reference construction, hybrid and dynamic short-read matching, big-data-guided dual-mode hierarchical taxonomy annotation strategy, and full-procedure parallel computing. This tool delivers species-level resolution and ultrafast speed for large-scale microbiomes, surpassing alignment-based approaches and the Naïve-Bayesian model. Furthermore, recognizing the global uneven distribution of microbes, we delineate optimal annotation strategies for each sampling habitat based on microbial patterns over 270,000 microbiomes. Integrated with the established workflow of Parallel-Meta Suite and the latest curated reference databases, this endeavor offers a swift and dependable solution for high-precision microbiome surveys.IMPORTANCEOur study introduces PM-profiler, a new tool that deciphers the complexity of microbial communities. With advanced algorithms, flexible annotation strategies, and well-organized big-data, PM-profiler provides a faster and more accurate way to study on microbiomes, paving the way for discoveries that could improve our understanding of microbiomes and their impact on the world.
摘要:
Amplicon测序是微生物组分析的基石,然而,人们仍然对其分辨率和准确性感到担忧。参考数据库和注释的增强标志着基于16SrRNA的谱分析的新时代。利用这种潜力,我们介绍PM-profiler,一种用于分析扩增子短读数的新工具。PM-profiler由基于C++的高级算法实现,例如用于参考构造的预分配哈希,混合和动态短读匹配,大数据引导的双模式分层分类注释策略,和全过程并行计算。该工具为大规模微生物组提供物种级分辨率和超快速度,超越基于对齐的方法和朴素贝叶斯模型。此外,认识到全球微生物分布不均,我们根据超过270,000个微生物组的微生物模式,为每个采样栖息地描绘最佳注释策略。与Parallel-MetaSuite的既定工作流程和最新策划的参考数据库集成,这一努力为高精度微生物组调查提供了快速可靠的解决方案。重要的是我们的研究介绍了PM-profiler,一种破译微生物群落复杂性的新工具。有了先进的算法,灵活的注释策略,和组织良好的大数据,PM-profiler提供了一种更快,更准确的方法来研究微生物组,为发现这些发现铺平道路,这些发现可以提高我们对微生物群及其对世界的影响的理解。
公众号