Mining the microbiome for markers of microbiota-gut brain communication and mental health

There has been a growing acknowledgement of the involvement of the gut microbiome - the collection of microbes that reside in our gut - in regulating our mood and behaviour. This phenomenon is referred to as the microbiota-gut-brain axis. While our techniques to measure the presence and abundance of these microbes has been steadily improving, there are many factors that prevent us from understanding what aspects of the gut microbiome specifically influence the microbiota-gut-brain axis. In this thesis, we set out to identify and investigate aspects of the microbiome that are informative to gut-brain communication. We do this by investigating the state of the gut microbiome in both health and disease, as well as after supplementing or perturbing it. While all of the work presented here is based on real data from real experiments, the thesis has a strong bioinformatics focus, that means that while the physiological background and interpretation are important, my role in these projects has been to bioinformatically and statistically zoom in on the features of the microbiome that are the most informative to our questions. As such, all results will be discussed from a primarily bioinformatics point of view. Two main aspects of the gut microbiome came out as the most promising features to measure, namely functional capacity and volatility. Traditionally, the microbiome is thought of as a collection of microbes and most analysis is done on the taxonomical level. However, we find that by investigating microbial function - as defined by the genes that are found or associated in the detected microbes - rather than taxonomy, we are able to perform more sensitive analysis and that our results are more easily interpretable. Second, microbiome studies are typically conducted using a single sample per subject. We find that the degree of change in the microbial ecosystem, called volatility, is an important feature of the microbiome and that is linked to severity of stress response. While volatility was coined before in the context of the microbiome, this was only in passing. We were the first to investigate volatility as a feature of the microbiome. Our research in this thesis reconfirms the existence of the microbiota-gut-brain axis and demonstrates novel metrics that can be used to interrogate the microbiome. We utilize mathematical frameworks originally from geology and classical ecology to bolster our analysis. We show that considering the microbiome as an ecosystem is a powerful model that can help us better formulate our scientific questions and interpret our findings. We argue for strategies to unify bioinformatics methodology in the microbiome-gut-brain axis field in an effort to move towards mechanistic understanding.