Vincent Rennie *1, Clare Warren 1,Sevi Filippidou 1, Kevin Purdy 2 & Karen Olsson-Francis 1
1 School of Environment, Earth and Ecosystem Sciences, Open University, Milton Keynes, MK7 6AA; vincent.rennie@open.ac.uk
2 School of Life Sciences, Gibbet Hill Campus, The University of Warwick, Coventry, CV4 7AL
At present, researchers are unable to grow the majority of microorganisms present on Earth in the laboratory. These microorganisms are often referred to as “microbial dark matter” and contribute to biogeochemical processes in natural systems, which through lack of laboratory investigation, remain unclear. The lack of understanding of how these organisms live remains a major obstacle for growing them in the laboratory. Recent genetic sequencing techniques have allowed researchers to generate hypotheses how microbial communities interact and function in diverse environments. However, to provide a better insight, we need to look at individual microorganisms; for this we need to grow them.
In my PhD, I have developed a method to verify the contributions of individual community members by linking functional potential with methods for isolating and growing them. Firstly, using machine learning (Figure 1B), I can confidently predict the broad trophic level of each community member in a multi-layered microbial community, collected from an acidic hydrothermal spring in the Azores (Figure 1A). Secondly, using pathway analysis software and data sorting, I can identify unique pathways associated with each community member. This information can be used to inform selective growth experiments for the successful isolation of these microorganisms.
Fig 1. (A) The biomat collected during the 2018 sampling excursion with layers numbered from pool to wall (B) The logical workflow of the machine learning methodology to determine the trophic level of each community member.
Several of the community members identified from the microbial community belong to groups of microorganisms without cultivated representatives. These findings highlight the potential for these predictive methodologies to develop genomics-informed growth and isolation studies. As the cost of high-throughput sequencing techniques continues to decrease, they can provide crucial community context information to help environmental microbiologists isolate the uncultivated majority.