Reimagining the Synthetic Biology DBTL Cycle with Machine Learning

Area of Science:

• Synthetic Biology
• Computational Biology
• Molecular Biology

Background:

• Synthetic regulatory circuit engineering is challenging due to intracellular complexity.
• Machine learning and artificial intelligence (ML/AI) show promise in accelerating circuit design.
• A key limitation in ML/AI for synthetic biology is the lack of high-quality training datasets.

Purpose of the Study:

• To detail methods for using the CLASSIC platform to build synthetic biology circuit libraries.
• To describe strategies for acquiring high-quality datasets for ML/AI model training.
• To highlight principles for integrating ML/AI into the synthetic biology design-build-test-learn cycle.

Main Methods:

• Utilized hierarchical DNA assembly to construct expansive circuit libraries.
• Employed next-generation long- and short-read sequencing for high-throughput data acquisition.
• Developed computational methods and strategies for data processing and ML/AI model training.

Main Results:

• The CLASSIC pipeline facilitates the generation of large-scale, high-quality datasets for synthetic biology.
• Established best practices for data collection and ML/AI model training in circuit engineering.
• Demonstrated the utility of CLASSIC in accelerating the design-build-test-learn cycle.

Conclusions:

• CLASSIC provides a robust platform for data-driven synthetic biology circuit design.
• The principles outlined can be generalized to other data-intensive projects in synthetic biology.
• ML/AI integration, powered by high-quality data, is crucial for overcoming complexity in synthetic biology.