A workflow for generating multi-strain genome-scale metabolic models of prokaryotes

Area of Science:

• Microbiology
• Systems Biology
• Computational Biology

Background:

• Genome-scale models (GEMs) are crucial for understanding bacterial metabolism.
• The increasing availability of genome sequences necessitates methods for multi-strain model generation.
• Existing protocols are limited in scalability for analyzing species-level metabolic diversity.

Purpose of the Study:

• To extend the original Protocol for generating single-strain GEMs to enable the creation of multi-strain GEMs.
• To provide a scalable and automatable procedure for constructing comprehensive metabolic models across bacterial strains within a species.
• To facilitate the study of pan-metabolic capabilities and strain-specific metabolic differences.

Main Methods:

• A four-stage extension of the original Protocol was developed.
• Stages include: reference strain model generation, genome sequence comparison for homology matrix creation, draft model generation, and manual curation.
• A supplementary Jupyter notebook tutorial aids in partial automation and scalability.

Main Results:

• The extended Protocol enables the generation of multi-strain GEMs.
• The method is scalable and can be partially automated, distinguishing it from the original Protocol.
• This approach allows for the investigation of pan-metabolic capabilities and strain-specific variations.

Conclusions:

• The developed Protocol Extension is a valuable tool for comparative metabolic modeling across bacterial strains.
• Multi-strain GEMs provide insights into a species' lifestyle range and evolutionary adaptations.
• This method complements existing model generation platforms like CarveMe and KBase.