Artificial intelligence: a solution to involution of design-build-test-learn cycle.

Area of Science:

• Biotechnology
• Synthetic Biology
• Computational Biology

Background:

• Iterative design-build-test-learn (DBTL) cycles are standard in microbial strain development, integrating computational design, genetic engineering, and omics analysis.
• These cycles can face involution, generating vast data and constructs without significant breakthroughs, hindering progress.
• Machine learning (ML) presents a potential solution for multiscale modeling and process optimization to overcome these limitations.

Purpose of the Study:

• To review recent advances in machine learning applications for microbial strain development.
• To highlight ML's role in integrative metabolic modeling and knowledge engineering.
• To discuss how ML can guide metabolic engineering and fermentation optimization for improved efficiency.

Main Methods:

• Review of current literature on ML applications in microbial strain engineering.
• Focus on ML for integrative metabolic models and knowledge engineering.
• Analysis of ML's potential in fermentation process optimization.

Main Results:

• ML offers a promising, yet underdeveloped, approach to multiscale modeling and process optimization in DBTL cycles.
• ML can guide metabolic engineering and fermentation optimization by integrating complex datasets.
• ML-based strain development can enhance the efficiency and effectiveness of DBTL cycles.

Conclusions:

• Machine learning can significantly improve microbial strain development by overcoming DBTL cycle involution.
• ML-powered approaches facilitate the transition of synthetic strains from laboratory to industrial scale.
• Further development and application of ML are crucial for advancing synthetic biology and biotechnology.