FastST: an efficient tool for inferring decomposition and directionality of microbial communities

FastST:一种用于推断微生物群落分解和方向性的有效工具

阅读:1

Abstract

Microbiomes play crucial roles in human health, disease development, and global ecosystem functioning. Understanding the origins, movements, and compositions of microbial communities is essential for unraveling the principles governing microbial ecology. Microbial source tracking (MST) approaches have emerged as valuable tools for quantifying the proportions of different microbial sources within target communities, enabling researchers to track transmissions between hosts and environments, identify similarities between microbiome samples, and determine sources of contamination in various settings. Current MST methods like SourceTracker2 and FEAST have advanced the field by employing Bayesian and expectation-maximization approaches, respectively, but are limited by computational inefficiency with high-dimensional data and inability to infer directionality in source-sink relationships. This study presents a novel computational framework for microbial source tracking called FastST. FastST infers the relative contributions of source environments to sink microbiomes while also determining directionality when source-sink relationships are not predefined. Through extensive simulation studies with varying numbers of sources and complexity, FastST demonstrates superior performance in both accuracy and computational efficiency compared to FEAST and SourceTracker2, maintaining consistent execution times even as the number of source environments increases. Furthermore, the proposed method achieved over 90% accuracy in directionality inference across all tested scenarios, even when multiple major sources are present, broadening its applicability in practical microbiome research and environmental monitoring. FastST and data simulation codes are publicly available at https://github.com/joungmin-choi/FastST.

特别声明

1、本页面内容包含部分的内容是基于公开信息的合理引用;引用内容仅为补充信息,不代表本站立场。

2、若认为本页面引用内容涉及侵权,请及时与本站联系,我们将第一时间处理。

3、其他媒体/个人如需使用本页面原创内容,需注明“来源:[生知库]”并获得授权;使用引用内容的,需自行联系原作者获得许可。

4、投稿及合作请联系:info@biocloudy.com。