Deciphering microbiome complexity using biological big data and artificial intelligence

Marine microalgae are diverse photosynthetic organisms with extensive biotechnological applications, such as bioremediation, generation of biofuels and biofeed, or as a pharmaceutical source. Typically microalgae cultivation systems are mono-cultures with protocols to prevent contamination, however recent research has shown beneficial effects of co-culturing with other microbes, e.g. higher lipid content for biofuel generation or faster and cheaper harvesting.

Still, little is known about community structures and interactions of microorganisms within the microalgal microbiome (phycosphere) and how they affect algal growth. Most current research concerns microalgae-bacteria systems, while few studies investigate microalgae-fungi communities, and even fewer consider interactions between all three. This project aims to identify culturable and non-culturable bacteria and fungi associated with the model marine microalga Nannochloropsis oceanica, develop efficient high-throughput methods to identify and map out interactions between isolated microorganisms, and test viability of using microalgal-associated microorganism co-cultures in biotechnological applications.

Microorganisms will be isolated from N. oceanica onto agar plates and identified using Sanger sequencing (culturable isolates) or Illumina NGS sequencing (non-culturable mixed communities) with 16S and ITS primer sets. High-throughput testing of microbial interactions will be done using an automated robot to pin together pairs of culturable colonies and the microalga onto solid agar plates.

Machine-learning image analysis of agar plate photographs will be implemented to assess whether the interactions are positive, negative, or neutral, based on surface area of colonies cultured alone or in a pair. Finally, the microalga will be co-cultured in solid and liquid media with selected microorganisms. The viability of using co-cultures in biotechnological applications (with a focus on bioremediation of endocrine disrupting chemicals) will be assessed by comparing factors such as the biomass, biochemical composition, flocculation rates, and removal efficiency of selected environmental contaminants, between co- and mono-cultures.

Results of this study will indicate whether this high-throughput methodology gives an accurate representation of interactions within the phycosphere and whether this can be utilised in biotechnology.