Human activities threaten the sustainability of Earth’s environment and require us to understand and predict ecosystem dynamics. However, ecosystems are often described using averaged indicators that fail to capture species-level differences and physiological states. Because key phenomena such as fishery resources and environmental responses depend on specific species, our laboratory aims to reconstruct ecosystems based on species-level functions and states.

We infer organismal functions and states from genomic and epigenomic data to characterize ecosystem structure and dynamics. In collaboration with the Onagawa Field Center and WPI-AIMEC, we study coastal plankton communities through field observations and sampling, integrating long-read metagenomic and epigenomic analyses. This approach captures species identity, abundance, and interactions, enabling description of community change. We also develop computational methods, including gene coexpression network analysis, and maintain databases (ATTED-II, COXPRESdb, ALCOdb). Together with PlanDyO, a metagenomic integration platform, these efforts enable description and prediction of ecosystem dynamics based on species-level processes.