Microbiome Season

Screen Shot 2017-05-09 at 21.25.41A central part of my diverse job at the University of Sydney is studying the gut microbiome, the trillions of microbes that inhabit our guts; how it influences our health, and how we can influence it through interventions such as diet. Aspects of this work are now, finally, getting into print.

This week my expert microbiologist collaborator (Dr. Andrew Holmes) and I published a Perspectives paper in Frontiers in Immunology. The paper explores the aspects of diet and our physiology that determine what sort of a microbial community can thrive within us. Simulation, modelling, bioinformatics and machine learning are advanced as essential tools in this endeavour. It’s not an easy system to study. Firstly, as a fellow human, you likely conjure a great displeasure at being “probed” in this regard, so studies of the system in its natural setting are problematic. Secondly, these microbes are very difficult to culture in the lab; our guts are lacking in oxygen, circumstances difficult (though not impossible) to reproduce in the real world. Further, our guts constitute a rich and complex environment of molecules that microbes can (and can’t) grow on, again difficult to reproduce exactly, as they are still the subject of study. Lastly, as exemplified by its analogy as a “fingerprint”, each of us harbours such a unique community of microbes that drawing parallels and similarities can be difficult. To meaningfully understand this community we need to go beyond (relatively) cheap and common sequencing techniques that log which individual microbes are present, and move into a space where we characterise their metabolic capacities and how they support and hamper one another’s growth when engaging in networks of cross feeding and interacting networks. This represents a move to functional analysis. And whilst we need to start collecting the thousands of distinct species of bacteria comprising our microbiomes into more manageable collections of similarly behaving organisms, to reduce the complexity of the analysis, the collections we assemble are problem- and response-specific. MeaninScreen Shot 2017-05-09 at 21.23.53g, no one size fits all. This is becoming a data-centric field of sequencing what we can, and then using machine learning and statistical analyses to spot the patterns. Ultimately, we’d like to predict an individual’s response to a given dietary intervention before they embark on a, say, weight-loss journey. Imagine if we could tell you what worked for you, the unique individual, as you tried to overcome irritable bowel syndrome, or some other of the many maladies in which the microbiome is implicated? Computational techniques are central to this field’s future.

Screen Shot 2017-05-09 at 21.22.41Our second success, first authored by the talented Dr. Andrew Holmes, is an ambitious diet intervention study that established our own body’s influence on the communities of microbes that are able to inhabit us. Published in the very prestigious Cell Metabolism journal, this study showed how the energy content per gram of food that is eaten balances off against the substances we excrete into our guts in shaping the microbial community. We are not inert habitats in this ecosystem; rather, the picture painted is one where we and our microbiome can be quite in tune with one another. Microbes can deliver to us 10% of our daily energy tally (if that doesn’t sound like much, contemplate a 10% pay cut and see how you feel about it) by fermenting fibres that we eat but can’t otherwise digest, and a host of vitamins and essential growth factors. In exchange for their occupying our guts as a niche, and thus preventing pathogens from gaining a foothold, we support these microbes during times of fasting by providing them with non-dietary, host-secreted, nutrients to grow in on the form of e.g. mucus (just recall the last time you had an aggressive cold). At least, that, we think, was the case classically, when securing a meal meant foraging or hunting. Now it means going to the supermarket, and we can eat a seemingly limitless quantity of nutrient rich food. Has this abundance of nutrient disrupted our traditional relationship with gut microbes? If there’s always dietary nutrient in our guts, then we are no longer inducing selective periods of dietary-nutrient-depletion that benefited only microbes with an evolutionary history-bestowed ability to grow on our supporting secretions? Rather, the range of microbes that can grow in our guts may have expanded to include microbes that don’t benefit us, and us them. Ok, this is all a bit speculative, and goes beyond what’s in the paper – but that’s why this is a blog, not a peer reviewed article. We can’t prove it with available data, but a lot of the signs are pointing in that direction. And being the microbiome, the situation is complicated. Finally, please remember, I’m a computer scientist, not a physician, dietician, microbiologist, or life coach. Whilst fasting in forms such as the 5-2 diet have become popular, the evidence of its benefit in humans is far from established, and in no way do I suggest you change your lifestyle based on what I’ve written here: go see an expert if you have concerns. My job in all of this is simulation and data analytics. And I love it.

Our work, my collaborators and I, has recently been covered in the University of Sydney alumni magazine.

There’s at least two more papers on this theme in the pipeline, hopefully both by the end of the year. Watch this space!