Secondary Metabolites Predict Diazotrophic Cyanobacteria: A Model-Based Cheminformatic Approach

Area of Science:

• Microbiology
• Biotechnology
• Metabolomics

Background:

• Nitrogen fixation (diazotrophy) is a key trait in cyanobacteria with agricultural and industrial potential.
• Identifying diazotrophic cyanobacteria is crucial for developing sustainable agricultural practices.
• This research explores the link between secondary metabolites and diazotrophy in cyanobacteria.

Purpose of the Study:

• To develop a predictive model for identifying diazotrophic cyanobacteria based on their secondary metabolites.
• To leverage chemical structure similarity for predicting diazotrophy.
• To prioritize potential diazotrophic strains from a large dataset of metabolites.

Main Methods:

• Developed a predictive algorithm using chemical fingerprint similarity of metabolites from CyanoMetDB.
• Evaluated the algorithm using leave-one-out cross-validation on 133 manually labeled metabolites.
• Applied the model to 1980 unlabeled metabolites to identify likely diazotrophic strains.

Main Results:

• The model achieved 88% accuracy and a ROC-AUC of 0.96 in predicting diazotrophy.
• Successfully prioritized potential diazotrophic strains among unlabeled metabolites.
• Toxicity analysis indicated metabolites are not defensive; high nitrogen and cyclic peptides suggest signaling roles.

Conclusions:

• Secondary metabolites can effectively identify diazotrophs, even without active diazotrophic physiology.
• This approach advances agricultural biotechnology by enabling the discovery of more diazotrophic cyanobacteria.
• Findings support the development of self-fertilizing crops and sustainable agriculture.