AI method reveals hidden patterns in microbial communities in the Warnow Estuary

A research team from the Leibniz Institute for Baltic Sea Research Warnemünde (IOW) has demonstrated that an AI method originally designed for text analysis can effectively identify patterns in complex microbial communities. The study, published in Environmental Microbiome, revealed five seasonally successive microbial subcommunities in the Warnow Estuary while maintaining ecological and functional information as accurately as conventional methods.

Microbial communities are highly sensitive to environmental changes, but their diversity makes it challenging to identify ecological patterns. The team led by Ph.D. student Anna Kujat applied topic modeling, a machine learning technique originally used for text analysis, to microbial communities. They applied this method to water samples collected from the Warnow Estuary and adjacent coastal waters of the southwestern Baltic Sea over a year-long period.

The data comprised 1,236 different bacterial genetic signatures, collected at 14 fixed stations. The researchers compared the topic modeling results with established methods, finding that it performed at least as well as, and sometimes better than, conventional methods. The results identified five microbial subcommunities characteristic of the Warnow Estuary, each with distinct taxonomic composition and environmental preferences.

These subcommunities appeared and disappeared in a seasonal sequence, reflecting the estuary's dynamic environmental conditions. The study highlights the potential of applying AI methods originally developed for text analysis to ecological research, providing a systematic and effective approach to analyzing complex environmental samples.