Multivariable association discovery in population-scale meta-omics studies.
Multivariable association discovery in population-scale meta-omics studies.
复制标题
DOI：
10.1371/journal.pcbi.1009442
复制
发表时间：
2021-11
影响因子：
4.3
通讯作者：
Huttenhower C
中科院分区：
文献类型：
--
作者：
Mallick H;Rahnavard A;McIver LJ;Ma S;Zhang Y;Nguyen LH;Tickle TL;Weingart G;Ren B;Schwager EH;Chatterjee S;Thompson KN;Wilkinson JE;Subramanian A;Lu Y;Waldron L;Paulson JN;Franzosa EA;Bravo HC;Huttenhower C
将人类健康结果、饮食、环境条件或其他元数据等特征与微生物群落测量结果相关联具有挑战性，部分原因是它们的定量特性。微生物组多组学通常具有噪声大、稀疏（零膨胀）、高维、极不正态的特点，并且通常以计数或成分测量的形式存在。在此，我们引入一种新方法和现有方法的优化组合，以在人群规模的观察性研究中评估微生物群落特征与复杂元数据的多变量关联。我们的方法MaAsLin 2（微生物组与线性模型的多变量关联）使用广义线性模型和混合模型，以适应各种各样的现代流行病学研究，包括横断面和纵向设计，以及各种有或没有协变量和重复测量的数据类型（例如计数和相对丰度）。为了构建这种方法，我们对一系列广泛的场景进行了大规模评估，在这些场景下，直接识别元组学关联可能具有挑战性。这些模拟研究表明，MaAsLin 2的线性模型在存在重复测量和多个协变量的情况下保留了统计功效，同时考虑了元组学特征的细微差别并控制了错误发现。我们还将MaAsLin 2应用于来自人类微生物组整合项目（HMP2）的微生物多组学数据集，除了重现已有的结果外，还揭示了炎症性肠病（IBD）在多个时间点和组学图谱上独特的综合情况。
最近，已经提出了几种统计方法来从微生物群落的分子图谱中识别与特征（例如分类群、基因、途径、化学物质等）的表型或环境关联。然而，特别是对于人类微生物组流行病学，这些方法中的大多数主要侧重于单变量关联，即仅分析一个或几个环境协变量。鉴于人群规模微生物组研究的日益普遍以及相关研究设计的复杂性，包括饮食、药物、临床和环境协变量，且通常有来自多个时间点或组织的样本，这是一个需要解决的关键差距。令人惊讶的是，对于适用于此类研究的统计分析方法没有系统的评估，对于可扩展的微生物组流行病学的适当方法也没有共识。为此，我们开发并验证了一种统计模型（MaAsLin），它提供了第一种统一的方法以及第一种在人群规模微生物群落研究中对多变量关联进行大规模、全面基准测试的方法。我们希望通过广泛模拟和应用于HMP2 IBD多组学而得到验证的MaAsLin 2的实施，将有助于研究人员未来对与人类相关和环境微生物群落的分析。
It is challenging to associate features such as human health outcomes, diet, environmental conditions, or other metadata to microbial community measurements, due in part to their quantitative properties. Microbiome multi-omics are typically noisy, sparse (zero-inflated), high-dimensional, extremely non-normal, and often in the form of count or compositional measurements. Here we introduce an optimized combination of novel and established methodology to assess multivariable association of microbial community features with complex metadata in population-scale observational studies. Our approach, MaAsLin 2 (Microbiome Multivariable Associations with Linear Models), uses generalized linear and mixed models to accommodate a wide variety of modern epidemiological studies, including cross-sectional and longitudinal designs, as well as a variety of data types (e.g., counts and relative abundances) with or without covariates and repeated measurements. To construct this method, we conducted a large-scale evaluation of a broad range of scenarios under which straightforward identification of meta-omics associations can be challenging. These simulation studies reveal that MaAsLin 2’s linear model preserves statistical power in the presence of repeated measures and multiple covariates, while accounting for the nuances of meta-omics features and controlling false discovery. We also applied MaAsLin 2 to a microbial multi-omics dataset from the Integrative Human Microbiome (HMP2) project which, in addition to reproducing established results, revealed a unique, integrated landscape of inflammatory bowel diseases (IBD) across multiple time points and omics profiles. Recently, several statistical methods have been proposed to identify phenotypic or environmental associations with features (e.g., taxa, genes, pathways, chemicals, etc.) from molecular profiles of microbial communities. Particularly for human microbiome epidemiology, however, most of these are primarily focused on univariable associations that analyze only one or a few environmental covariates. This is a critical gap to address, given the growing commonality of population-scale microbiome research and the complexity of associated study designs, including dietary, pharmaceutical, clinical, and environmental covariates, often with samples from multiple time points or tissues. Surprisingly, there have been no systematic evaluations of statistical analysis methods appropriate for such studies, nor consensus on appropriate methods for scalable microbiome epidemiology. To this end, we developed and validated a statistical model (MaAsLin) that provides both the first unified method and the first large-scale, comprehensive benchmarking of multivariable associations in population-scale microbial community studies. We hope that the MaAsLin 2 implementation, validated through extensive simulations and an application to HMP2 IBD multi-omics, will be helpful for researchers in future analysis of both human-associated and environmental microbial communities.
登录
查看更多内容
影响因子：
12.3
作者：
Mallick H;Ma S;Franzosa EA;Vatanen T;Morgan XC;Huttenhower C
通讯作者：
Huttenhower C
影响因子：
9.5
作者：
Hawinkel, Stijn;Mattiello, Federico;Thas, Olivier
通讯作者：
Thas, Olivier
影响因子：
--
作者：
Mandal S;Van Treuren W;White RA;Eggesbø M;Knight R;Peddada SD
通讯作者：
Peddada SD
DOI：
10.1111/j.2517-6161.1995.tb02031.x
发表时间：
1995-01-01
影响因子：
5.8
作者：
BENJAMINI, Y;HOCHBERG, Y
通讯作者：
HOCHBERG, Y
DOI：
10.1038/nrmicro3451
发表时间：
2015-06
期刊：
Nature reviews. Microbiology
影响因子：
--
作者：
Franzosa EA;Hsu T;Sirota-Madi A;Shafquat A;Abu-Ali G;Morgan XC;Huttenhower C
通讯作者：
Huttenhower C