Combining Metagenomics with Structural Bioinformatics Reveals the Selective Pressures Driving Protein Evolution in Globally-prevalent Microbial Populations

Microbes play important roles in disease, human health, and climate change. Understanding how environmental selective forces shape their evolution underpins our ability to prevent, promote, and engineer their behavior. The genetic diversity of microbial populations can be quantified with metagenomics, however, such diversity represents the outcome of both stochastic and selective forces, making it difficult to identify whether variants are maintained by adaptive, neutral, or purifying processes. This is partly due to the reliance on gene sequences to interpret variants, which disregards the physical properties of three-dimensional gene products that define the functional landscape on which selection acts. Although it is understood that the accuracy of sequence-based evolutionary models improves by integrating structural information of the encoded protein, including structural bioinformatics into metagenomic analyses is hampered by the absence of computational tools that allow researchers to seamlessly integrate these traditionally distinct data types. In my dissertation, I bridge this disconnect by developing anvi'o structure, a computational tool for the analysis and visualization of metagenomic sequence variants in the context of predicted protein structures and binding sites. Taking a marine microbial population as a model system, I illustrate how structure-informed analyses yield insight into the evolutionary relationship between microbes and their environments that can only be learned by combining metagenomics with structural biology. Overall, my work sheds light on how environments induce selective pressures that in turn impact the genetic diversity of populations, and provides a software tool that enables the community to employ similar analyses on different microbial systems.