DyMMM-LEAPS: An ML-based framework for modulating evenness and stability in synthetic microbial communities

Area of Science:

• Synthetic biology
• Computational biology
• Microbial ecology

Background:

• Designing stable synthetic microbial consortia requires computational tools to navigate complex parameter spaces.
• Maximizing community properties like evenness and stability is crucial for reliable synthetic microbial systems.

Purpose of the Study:

• Introduce DyMMM-LEAPS (dynamic multispecies metabolic modeling-locating evenness and stability in large parametric space), a novel framework for synthetic microbial community design.
• Identify regions in the genetic circuit parameter space that promote high evenness and stability in microbial consortia.

Main Methods:

• Utilize adaptive sampling and surrogate modeling to overcome computational costs associated with exhaustive parameter space exploration.
• Extend the dynamic multispecies metabolic modeling (DyMMM) framework to analyze large parametric spaces.
• Simulate cocultures and polycultures with varying social interactions (cooperation, competition, predation) using quorum-sensing-based genetic circuits.

Main Results:

• DyMMM-LEAPS successfully predicts engineering targets and their operating ranges for enhanced evenness and stability.
• Demonstrated the framework's efficacy on multiple synthetic microbial community configurations, including cocultures and a three-strain culture.
• The analysis revealed specific 'pockets' of evenness and stability, offering insights into the relationship between these properties.

Conclusions:

• DyMMM-LEAPS provides a computationally efficient method for designing robust synthetic microbial communities.
• The framework aids in tuning genetic circuits and dissecting the interplay between community evenness and stability.
• DyMMM-LEAPS is adaptable for larger, more complex microbial communities and interactions.