A Computational Framework for Designing Microbiome Manipulation

描述(由申请人提供) 摘要：人体微生物群--人体内的复杂微生物群--对我们的健康有巨大影响。世界范围内的研究计划现在提供了对微生物组以前未知的组成的初步见解，并揭示了与广泛的疾病和宿主表型相关的显著成分变化。尤其是肠道微生物组，在许多必需的过程中起着关键作用，并积极促进各种疾病的发生，包括肥胖、糖尿病、炎症性肠病、心血管疾病和神经系统疾病。具体地说，最近的研究表明，将供者微生物区系转移到受者体内可以诱导不同的供者表型或促进患病受者的康复。这些发现表明，通过直接操纵肠道微生物组是一种有希望的治疗途径。例如，这种临床干预可以针对其组成代表某种疾病状态的微生物组，如糖尿病，并促进组成向健康配置的转变。或者，可以用一种新的具有某些首选代谢能力的“设计者”微生物群来定居个人，例如，允许营养不良的儿童群体从有限的饮食中获得更多的能量。然而，为了使基于微生物组的治疗向前发展并充分发挥其潜力，需要一个全面的计算工具包来指导这种操作并提出有希望的干预路线。在这个项目中，我们将相应地开发一个计算框架，以设计针对一组特定的所需代谢目标的微生物组操作。这一框架将包括两个主要组成部分：系统级微生物组代谢的计算机内模型，能够成功地预测给定微生物群在肠道环境中的代谢活动；以及优化模块，将用于搜索可能的微生物组组成空间，以寻找与所需代谢目标最接近的微生物组。我们将使用多种代谢建模和分析方法，并提出新的计算方法来研究复杂的多物种群落。这种系统级计算方法在研究和设计单一物种新陈代谢方面被证明是极其强大和有效的，但尚未被应用于研究社区范围的新陈代谢。还将开发计算技术，以考虑不同物种的丰度，并评估设计的微生物群的弹性。这项拟议的跨学科研究将计算系统生物学、计算机内模型和复杂网络分析与基因组和元基因组数据相结合，在构建指导和告知基于微生物组的临床干预的计算框架方面迈出了关键的第一步。该项目代表着在研究人类微生物组和提供计算工具方面的重大飞跃，以直接利用这一新获得的知识促进人类健康。 公共卫生相关性：人类微生物群对我们的健康有巨大的影响，并与从肥胖和糖尿病到心血管疾病和神经紊乱的各种疾病状态有关。微生物组操作，无论是通过靶向的特定干预，还是通过整个微生物组移植，都是一个令人兴奋的临床前沿，具有许多有前途的医学应用。在这个项目中，我们将开发一个新的计算框架，将微生物组及其对宿主的影响的预测模型与优化技术相结合，用于设计此类操作并为临床干预工作提供信息。

DESCRIPTION (Provided by the applicant) Abstract: The human microbiome - the complex ensemble of microorganisms that populate the human body - has a tremendous impact on our health. World-wide research initiatives now provide preliminary insights into the previously uncharted composition of the microbiome, and reveal marked compositional changes associated with a wide range of diseases and host phenotypes. The gut microbiome, especially, plays a key role in many essential processes and actively contributes to various disease states, including obesity, diabetes, inflammatory bowel disease, cardiovascular diseases, and neurological disorders. Specifically, recent studies have demonstrated that transferring a donor microbiota into a recipient can induce various donor phenotypes or prompt the recovery of a sick recipient. These findings suggest a promising therapeutic avenue via directed manipulation of the gut microbiome. Such clinical interventions could target, for example, a microbiome whose composition typifies a certain disease state, such as diabetes, and promote a compositional shift into a healthy configuration. Alternatively, individuals could be colonized with a new ""designer"" microbiome with some preferred metabolic capacities, allowing, for example, populations of undernourished children to harvest more energy from a limited diet. However, to allow microbiome-based therapy to move forward and to realize its full potential, a comprehensive computational toolkit is required for directing such manipulations and proposing promising intervention routes. In this project, we will accordingly develop a computational framework for designing microbiome manipulation targeted at a specific set of desired metabolic goals. This framework will encompass two main components: A system-level in-silico model of microbiome metabolism, capable of successfully predicting the metabolic activity of a given microbial consortia in the gut environment, and an optimization module which will be used to search the space of possible microbiome compositions for those that most closely match the required metabolic goals. We will use multiple metabolic modeling and analysis approaches, and present novel computational methods for studying complex multi-species communities. Such system-level computational methods proved extremely powerful and effective in studying and engineering single species metabolism, but have not yet been applied to study community-wide metabolism. Computational techniques will also be developed to account for varying species abundances and to assess the resilience of designed microbiomes. This proposed cross-disciplinary research integrates computational systems biology, in-silico models, and complex networks analysis, with genomic and metagenomic data, taking a first crucial step in the construction of a computational framework for guiding and informing microbiome-based clinical interventions. This project represents a major leap forward in the study of the human microbiome and in providing computational tools for harnessing this newly gained knowledge directly for promoting human health. Public Health Relevance: The human microbiome has a tremendous impact on our health and has been associated with various disease states ranging from obesity and diabetes to cardiovascular diseases and neurological disorders. Microbiome manipulation, either via targeted specific intervention or via whole microbiome transplantation, is an exciting clinical frontier with numerous promising medical applications. In this project, we will develop a novel computational framework, integrating a predictive model of the microbiome and its impact on the host with optimization techniques, for designing such manipulations and for informing clinical intervention efforts.