Investigating Diet and Gut Microbiota Composition in an Obese Population using Shotgun Metagenomic Sequencing (InDiGO)

人类肠道菌群的组成与大量疾病状态有关，尤其是肥胖。与此同时，饮食是影响肠道微生物群的可改变因素之一。因此，已经投入了大量资源来了解人类肠道微生物群在人类健康中可能发挥的作用。然而，一些基本问题仍未得到解决，例如如何在微生物组研究中最好地描述饮食摄入（食物组、营养素组、饮食模式），长期饮食在多大程度上与微生物群组成相关，这种关系对人类健康有何意义，如何定义“健康”的微生物组，哪些微生物类群或微生物亚群与肥胖相关，以及在多大程度上功能分析、血液和尿液代谢组学数据，高分辨率的分类鉴定在微生物组研究中很有用。在本提案中，我们的目标是使用MeGA（代谢健康奥格斯堡）研究的数据来解决这些问题，这是一项238名肥胖或体重正常的成年人的研究。除了大量其他参数外，在9个月的研究期间，还收集了粪便样本的代谢组学数据、基于重复24小时食物清单和食物频率问卷的高质量饮食数据以及全基因组霰弹枪深度宏基因组测序数据。因此，MeGA研究提供了一个很好的机会，可以继续我们之前使用机器学习识别人类微生物组中潜在微生物亚群的研究（任务2.1）。在任务3.1-3.3中，我们还将评估微生物群组成（在物种和菌株水平上）、功能分析和代谢组学数据与习惯性饮食之间的关系。在这些分析中，习惯性饮食将在营养、食物组和饮食模式水平上进行分析。MeGA研究提供了在研究人群中进行复制的独特机会。因此，所有主要分析将在第一次样本后9个月提供第二次粪便样本的所有参与者中重复进行（n=200）。此外，所有重要的结果将在肥胖受试者的敏感性分析中得到重复，以评估体重对这些关联的影响。

The composition of the human gut microbiota is associated with a huge number of disease states, including obesity in particular. At the same time, diet is among the modifiable factors that affect the gut microbiome. As a result, substantial resources have been invested into understanding the role the human gut microbiome may play in human health. However, several basic questions remain unanswered, such as how to best describe dietary intake in microbiome studies (food groups vs. nutrients vs. dietary patterns), to what extent long-term diet is associated with microbiota composition and what significance this relationship has to human health, how to define a “healthy” microbiome, which microbial taxa or microbial subgroups are associated with obesity, and to what extent functional profiling, blood and urine metabolomics data, and high-resolution taxonomic identification is useful in microbiome studies. In this proposal, we aim to address these questions using data from the MeGA (Metabolic Health Augsburg) study, a well-characterized study population of 238 adults who are either obese or normal-weight. In addition to a host of other parameters, metabolomics data, high-quality dietary data based on repeated 24-hour food lists and a food frequency questionnaire, and deep whole genome shotgun metagenomic sequencing data in stool samples were gathered over the 9-month study period. Thus, the MeGA study provides an excellent opportunity to continue our previous research on identifying latent microbial subgroups within the human microbiome using machine learning (Task 2.1). We will also assess the relationship between microbiota composition (at the species- and, where possible, strain-level), functional profiling, and metabolomics data with habitual diet in Tasks 3.1-3.3. In these analyses, habitual diet will be analyzed at the nutrient, food group, and dietary pattern level. The MeGA study presents a unique opportunity for replication within the study population. Therefore, all main analyses will be replicated with all participants who provided a second stool sample nine months after the first sample (n=200). Additionally, all significant results will be replicated in a sensitivity analysis in obese subjects to evaluate the impact of body weight on the associations.