Assembly Rules of the Microbiome

Thesis (Ph.D.)--University of Washington, 2015The microbiome - the communal ecosystem of bacteria, archaea, and microscopic eukarya as well as the environments sustaining them - surpasses its macroscopic counterpart in taxonomic diversity, metabolic capabilities, and biogeographical distribution. Microbes inhabit every habitat hospitable to life, even those too extreme for higher order organisms. In milder environments, microbes provide the foundation for all life, either by maintaining essential biogeochemical cycles or by directly supporting host macroorganism health. These microbes interact to form complex communities, some of which contain up to thousands of distinct species. Modern metagenomic techniques have revealed extensive structure in these communities in the form of non-random species co-occurrence patterns. It has been proposed that patterns such as these arise through cogent assembly rules - deterministic ecological processes that govern the characteristic structure of a community. Yet, because the physiology and behavior of only a tiny fraction of microorganisms has been characterized, it becomes challenging to distinguish between alternative sets of community assembly rules. In this dissertation, I describe work l performed during my doctoral studies to develop analytical frameworks that integrate community metagenome information with single species whole genome information to identify community assembly rules that structure the human and global microbiomes. In chapter 1 I discuss the context for this work: the microbial communities studied, the methods used to characterize them, and the specific challenges I aim to address. In chapter 2 I describe how I used metabolic models of species interaction to determine that habitat filtering, and not species assortment, determines the structure of the human microbiome. These models additionally showed that host intestinal health state does not define the axes along which the community is filtered, hinting at more subtle biochemical processes yet to be determined. In chapter 3 I describe an analysis of functional complementarity, interactions where microbes bring to a shared environment a pair of functions not typically encoded within a single genome. Investigating this interaction on a large scale across a network of co-occurring microbes, I demonstrate that niche partitioning, rather than cooperative interaction, structures the global microbiome. In chapter 4 I describe a comparison of genetic co-occurrence across genomes and metagenomes. Differential analysis of co-occurrence structure reveals that while metagenome structure resembles that found in genomes, processes such as environmental remediation are still likely to be distributed across community members. Finally, in chapter 5 I discuss how these particular assembly rules relate to one another at different ecological scales, and offer perspective as to the future development and application of the analytical frameworks presented here