Collaborative Research: DMS/NIGMS 2: New statistical methods, theory, and software for microbiome data

高通量测序技术的进步使微生物组的特征能够通过 标记基因(例如，16S rRNA基因)扩增子测序或元基因组鸟枪法测序。 因此，科学界越来越认识到 微生物群落在许多人类健康和疾病状况中发挥作用。尽管很受欢迎，但 然而，微生物组和元基因组学研究领域还没有达到#年所达到的成熟程度。 其他已建立的分子流行病学领域，如癌症生物标记物的发现和全基因组 从组学研究向合理微生物组疗法跨越的联合研究。 利用大量微生物组和元基因组学数据的主要限制之一是 计算和统计方面的挑战。其中包括数据的技术性质，包括高 维度、稀疏计数或组合数据结构、相对较小的样本大小和复杂 依赖/关联结构，如系统发育关系。为了应对这些挑战，这 该提案寻求开发统计方法、理论和计算工具来准确地描述 大型研究内部和跨大型研究的微生物群落，同时保持统计严谨性和 生物相关性。这个项目开发了新的统计方法、理论和软件来表征 在大型研究中和在大型研究中准确地研究微生物群落。具体地说，是由生物医学推动的 以及在皮肤病、自闭症谱系障碍的微生物组研究中遇到的生物学问题， 和婴儿生长，研究人员将开发统计方法来(1)绘制微生物分类群图， 以强大而稳健的模式影响感兴趣的临床结果；(2)学习相关性 微生物类群之间的结构，以解码微生物群之间的复杂网络和相互作用 (3)高维微生物谱微生物组研究的一种新中介分析 以及代谢组学等其他组学特征。这项提案的成功完成将填补这一空白 在微生物组研究的蓬勃发展的研究兴趣和对更多分析工具的需求之间。 这一建议将提高对许多健康的潜在微生物组机制的理解 和疾病状况，这是设计基于微生物组的干预措施以预测预后的关键， 用于诊断和治疗目的。

Advancement in high-throughput sequencing technology allows the characterization of the microbiome via either marker-gene (e.g., 16S rRNA gene) amplicon sequencing or metagenomics shotgun sequencing. Consequently, the scientific community is increasingly appreciative of the important role that the microbiome community plays in many human health and disease conditions. Despite its popularity, the field of microbiome and metagenomics studies, however, has not yet reached the maturity attained in other established molecular epidemiology fields, such as cancer biomarker discovery and genome-wide association studies for making the leap from omics survey to rational microbiome-based therapeutics. One of the primary limitations to leveraging this large body of microbiome and metagenomics data is computational and statistical challenges. Among these is the technical nature of the data, including high dimensionality, sparse count or compositional data structure, relatively small sample size, and complex dependence/correlation structure such as phylogenetic relatedness. To combat these challenges, this proposal seeks to develop statistical methods, theory, and computational tools to accurately characterize microbial communities within and across large studies while maintaining both statistical rigor and biological relevance. This project develops new statistical methods, theory, and software to characterize microbial communities within and across large studies accurately. Specifically, motivated by biomedical and biological problems encountered in microbiome studies of skin diseases, autism spectrum disorder, and infant growth, the investigators will develop statistical methodology for (1) mapping microbial taxa that influence clinical outcomes of interest in a powerful and robust pattern; (2) learning the correlation structure among microbial taxa to decode the complex networks and interactions among the microbiome community; (3) a new mediation analysis for microbiome studies with high-dimensional microbial profiles and other omics profiles such as metabolomics. Successful completion of this proposal will fill the gap between the burgeoning research interests in microbiome studies and the need for more analytical tools. This proposal will improve the understanding of the underlying microbiome mechanism of many health and disease conditions, which is critical to designing microbiome-based interventions for prognostic, diagnostic, and treatment purposes.