Mathematical ecology models of host-microbiota interaction in auto microbiota transplants (auto-FMT)

项目摘要 寄主-微生物区系相互作用的数学生态学模型 在自动微生物区系移植中(AUTO-FMT) 我们的目标是开发合理设计微生物区系移植的数学模型，以恢复 抗生素治疗患者受损微生物区系的组成多样性和功能。我们将重点关注 接受异基因造血干细胞移植(allo-HSCT)的住院癌症患者。ALLO-HSCT 是一种潜在的治愈癌症的方法，它会损害免疫系统，并要求患者 接受大规模的抗生素治疗，以预防和治疗危及生命的感染。我们将在巨大的基础上 临床数据库、生物反应器体外实验和小鼠体内实验发展动态 描述抗生素如何引起微生物组成变化的数学模型，以及如何 会影响异基因造血干细胞移植后宿主免疫系统的恢复。该模型扩展了方法 由我们的团队首创的广义Lotka Volterra生态回归(GLOVER)和基于代理的 模型-走向一种可以帮助患者开发微生物区系疗法的模型 Allo-HSCT。 在目标1中，我们将使用斯隆·凯特林纪念馆提供的独特临床资源的数据 癌症中心-从1500名allo-HSCT患者获得的样本库(包括微生物组16S rRNA和 鸟枪测序)和广泛的临床元数据(包括完整血细胞计数和时间的时间序列 以及在患者住院期间给予的所有药物的剂量)；我们还将使用来自首个此类数据 自体粪便微生物区系移植在异基因造血干细胞移植中的随机对照试验 病人。我们将使用这些独特的资源来参数化我们的模型，并研究微生物区系如何 成分影响宿主免疫系统的恢复。在目标2中，我们将验证微生物 我们数学模型的组成部分，使用来自厌氧实验室反应器的实验数据 在缺乏抗生素治疗和自身FMT的情况下，在体外重建人类微生物区系的动态 一个活生生的主人。在目标3中，我们将开发小鼠模型来研究相同的微生物区系动态 在实验中，但现在是在一个活的宿主的背景下。 从这些临床研究、体外实验和体内模型中获得的数据将完善我们的 数学模型在模拟和定量实验的封闭循环中。我们的最终目标是 开发可以定义最佳微生物鸡尾酒并重建异种微生物区系的模型 HSCT患者。在这个过程中，我们希望为未来的治疗揭示微生物区系生态学的一般原理。 在微生物区系因抗生素治疗而受损的其他患者群体中。

Project Summary Mathematical ecology models of host-microbiota interaction in auto microbiota transplants (auto-FMT) We aim to develop mathematical models for the rational design of microbiota transplants that can restore compositional diversity and function to the damaged microbiota of antibiotic-treated patients. We will focus on hospitalized cancer patients receiving allogeneic hematopoietic stem cell transplants (allo-HSCT). Allo-HSCT is a potentially curative cancer treatment that compromises the immune system, and requires that patients receive massive antibiotic treatments to prevent and treat life-threatening infections. We will build on a vast clinical database, in vitro experiments in bioreactors and in vivo experiments with mice to develop dynamic mathematical models that describe how antibiotics cause changes in the microbial composition, and how that can impact the recovery of the host's immune system after allo-HSCT. The model expands approaches pioneered by our team—the Generalized Lotka Volterra Ecological Regression (GLOVER) and agent-based models—towards a model that can assist in the development of microbiota therapies for patients undergoing allo-HSCT. In aim 1 we will use data from a unique clinical resource available at the Memorial Sloan Kettering Cancer center—a sample bank obtained from >1,500 allo-HSCT patients (including microbiome 16S rRNA and shotgun sequencing) and extensive clinical metadata (including time series of complete blood counts and time and doses of all drugs given while the patients are hospitalized); we will also use data from a first-of-its-kind controlled randomized trial of autologous fecal microbiota transplant (auto-FMT) undergoing in allo-HSCT patients. We will use these unique resources to parameterize our models and investigate how the microbiota composition influences the recovery of the host immune system. In aim 2 we will validate the microbial component of our mathematical model using experimental data from anaerobic laboratory reactors that recreate—in vitro—the human microbiota dynamics during antibiotic treatment and auto-FMT in the absence of a living host. In aim 3 we will develop mouse models to investigate those same microbiota dynamics experimentally but now in the context of a living host. The data obtained from these clinical studies, in vitro experiments and in vivo models will refine our mathematical models in close cycles of simulation and quantitative experimentation. Our ultimate goal is to develop models that can define optimal microbial cocktails and reconstitute the perturbed microbiota of allo- HSCT patients. In the process we hope to uncover general principles of microbiota ecology for future therapies in other patient populations whose microbiota is damaged by antibiotic treatments.