Integrating metagenomics and geochemistry: functional evolution and taxonomic classification of hot spring communities

The taxonomic and metabolic profile of the microbial community inhabiting a natural system is largely determined by the physical and geochemical properties of the system. However, the influences of parameters beyond temperature, pH and salinity have been poorly analyzed with few studies incorporating the comprehensive suite of physical and geochemical measurements required to fully investigate the complex interactions known to exist between biology and the environment. Further, the techniques used to classify the taxonomic and functional composition of a microbial community are fragmented and unwieldy, resulting in unnecessarily complex and often non-consilient results.

This dissertation integrates environmental metagenomes with extensive geochemical metadata for the development and application of multidimensional biogeochemical metrics. Analysis techniques including a Markov cluster-based evolutionary distance between whole communities, oligonucleotide signature-based taxonomic binning and principal component analysis of geochemical parameters allow for the determination of correlations between microbial community dynamics and environmental parameters. Together, these techniques allow for the taxonomic classification and functional analysis of the evolution of hot spring communities. Further, these techniques provide insight into specific geochemistry-biology interactions which enable targeted analyses of community taxonomic and functional diversity. Finally, analysis of synonymous substitution rates among physically separated microbial communities provides insights into microbial dispersion patterns and the roles of environmental geochemistry and community metabolism on DNA transfer among hot spring communities.

The data presented here confirms temperature and pH as the primary factors shaping the evolutionary trajectories of microbial communities. However, the integration of extensive geochemical metadata reveals new links between geochemical parameters and the distribution and functional diversification of communities. Further, an overall geochemical gradient (from multivariate analyses) between natural systems provides one of the most complete predictions of microbial community functional composition and inter-community DNA transfer rates. Finally, the taxonomic classification and clustering techniques developed within this dissertation will facilitate future genomic and metagenomic studies through enhanced community profiling obtainable via Markov clustering, longer oligonucleotide signatures and insight into PCR primer biases.