BEACON Researchers at Work: Linking microbial interactions to disease

This week’s BEACON Researchers at Work blog post is by University of Idaho graduate student Daniel Beck. 

I am interested in microbial communities for a number of reasons. First, microbial communities are found nearly everywhere, from soil to the surfaces of everyday objects. Moreover they interact profoundly with other organisms, even establishing complex communities on the skin and in the guts of animals. Additionally, microbial communities do important things such as playing an important role in nutrient cycling in the soil. They can even affect human health, for example aiding in digestion and playing an important role in disease.

However, while microbial communities are common, they are difficult to study. There are huge numbers of individual microbes in a community and microbes are very diverse. The number of microbial cells in a single gram of soil is estimated to be on the order of 10^8. On top of that, hundreds or thousands of different microbe types may inhabit a single community. But this complexity is only the first hurdle to studying microbial communities. We are not yet able to grow many microbes in the lab. Even if such growth was possible, microbes may behave differently when grown in isolation compared to when they are a part of a complex community.

Luckily, new sequencing technologies are allowing us to study microbial communities in new and dynamic ways. Specifically, the development of two basic techniques allows us to accommodate many of these challenges. Bar code DNA sequencing lets us estimate the relative number of each type of microbe in the community. Similarly, metagenomic sequencing can show which genes are present in the community, allowing us to detect potential biochemical pathways. These data are giving us a new view of microbial communities. It also allows us to study the microbial communities without needing to grow the microbes in the lab. These techniques have made it possible to research microbial communities in new ways.

My research is focused on microbial communities that may be related to bacterial vaginosis (BV). BV is a disease characterized by changes in the microbial community in the vagina. It is remarkably prevalent, with estimates as high as nearly 30% of all women affected. Moreover BV is particularly interesting because it appears to be linked to the type of microbial community, rather than individual types of microbes.

Specifically my research is to determine which parts of the microbial community in the vagina are associated with BV. If BV is not linked to a specific microbe, perhaps it is linked to interactions between the microbes. Moreover, interactions between the microbes and environmental factors may be important. Researchers studying genetic interactions have developed a number of tools that aim to detect interactions between genes that may lead to disease, which is remarkably similar to the search for microbial interactions. Because I wish to detect interactions between microbes that are associated with BV, I am applying many of the same tools as researchers studying genetic interactions. Some of these tools include logistic regression, multifactor dimensionality reduction, and genetic programming.

All of these techniques work in similar ways. First, models are generated that relate aspects of the microbial community to BV. These aspects may be the relative abundance of microbe types or various environmental factors. After these models are generated, I evaluate them based on how well they are able to classify microbial communities into BV and non-BV categories. I then analyze the models to determine which aspects of the microbial community are being strongly linked to the classification. The parts of the community identified by the models as important become strong candidates for further study.

But this is just the beginning! In particular, I hope to extend these tools to new types of data. If these techniques work to detect microbial interactions related to BV, then they may also be able to detect interactions in other types of microbial communities. In addition, these same tools may be applied to more detailed metagenomic data to look for interactions between genes in microbial communities.

It is an exciting time to be a graduate student studying microbial communities. It often seems like every answer leads to many more questions about how microbial communities work. New types of data and analysis techniques are allowing us slowly begin to understand the incredibly complex world of microbes.

For more information about Daniel’s work, you can contact him at danlbek at gmail dot com.

This entry was posted in BEACON Researchers at Work and tagged , , , . Bookmark the permalink.

Comments are closed.