Predicting function from taxonomy to reveal insights into the human microbiome

Bacteria are living all over and within our bodies, and in fact outnumber our own cells 10 to 1. These are mainly "good" bacteria that reside in normal healthy people, but can be killed by the same antibiotics that are used to treat us when we are sick with "bad" bacteria. Until recently, not much was known about the different species of bacteria living on us and how that differed across different sites on our bodies, between individuals, and over our lifetime. Interestingly, recent research has shown that these communities likely have an impact on our nutrition, obesity, immune system, and various conditions such as Crohn's disease and Irritable Bowel Syndrome. The method used to discover these relationships is to identify the bacterial species and molecular functions that those species encode across various types of people. This is done by sequencing the DNA of each sample as a single jumbled bag of random fragments. Highly sophisticated computer programs are then used to figure out which fragments likely came from a particular species and what bacterial genes are associated with different types of people. My proposed research is to develop new computational algorithms to better highlight the biological interesting differences between microbiomes from varying sites and people. My methods will utilize the massive amounts of information contained within the more than 2000 bacterial genomes that have already been sequenced by the research community. This information will be used to predict the functions encoded by bacteria that we have not sequenced yet, but are present within our mircobiomes. Once these computer programs are implemented, they will be freely available for all researchers to use and will allow us as well as others to obtain further insight into the human microbiome and its effect on our health.