From mechanism to application: Decrypting light-regulated denitrifying microbiome through geometric deep learning

Regulation on denitrifying microbiomes is crucial for sustainable industrial biotechnology and ecological nitrogen cycling. The holistic genetic profiles of microbiomes can be provided by meta-omics. However, precise decryption and further applications of highly complex microbiomes and corresponding meta-omics data sets remain great challenges. Here, we combined optogenetics and geometric deep learning to form a discover-model-learn-advance (DMLA) cycle for denitrification microbiome encryption and regulation. Graph neural networks (GNNs) exhibited superior performance in integrating biological knowledge and identifying coexpression gene panels, which could be utilized to predict unknown phenotypes, elucidate molecular biology mechanisms, and advance biotechnologies. Through the DMLA cycle, we discovered the wavelength-divergent secretion system and nitrate-superoxide coregulation, realizing increasing extracellular protein production by 83.8% and facilitating nitrate removal with 99.9% enhancement. Our study showcased the potential of GNNs-empowered optogenetic approaches for regulating denitrification and accelerating the mechanistic discovery of microbiomes for in-depth research and versatile applications. This study proposed and showcased the discover-model-learn-advance (DMLA) cycle for unlocking nature as codebases to empower biological mechanisms decryption and biotechnology development. At the discovering stage, we found the bidirectionally light-regulated denitrification. Then, metatranscriptomic was utilized for geometric deep learning to obtain coexpressed gene panels. On the basis of that, we developed toolkits to learn underlying mechanisms and advance biotechnology, which were validated in the wet lab. Deepening recognition can further drive continuous DMLA cycles to accelerate scientific and biotechnology development.image Graph neural networks (GNNs)-based biology-contextualized computational framework exhibited superior performance in identifying coexpressed gene panels and decrypting wavelength-dependent denitrification.Wet-lab demonstrations validated the wavelength-divergent secretion system and nitrate-superoxide co-regulation as unveiled by GNNs, which could be utilized for nitrate removal and resource recovery.The coexpressed gene panels and topological network toolkits were developed to guide scientific discovery and versatile biotechnology development.