Machine learning-assisted synthetic biology of cyanobacteria and microalgae

Abstract

Cyanobacteria and microalgae have recently been utilized as autotrophic chassis for microbial cell factories capable of converting light and CO₂ directly into high-value products. However, compared to other heterotrophic model chassis, they face many challenges such as limited genetic manipulation tools, poorly understood metabolic regulatory networks, and low production efficiency. Machine learning, a data-driven pattern recognition strategy, leverages statistical associations in datasets for classification and generation, proving to be a valuable tool for synthetic biology. Synthetic biology requires the extraction of fundamental component information from complex biological data to create novel applications through the reassembly of biological components. This process aligns well with the capabilities of machine learning. In this review, we briefly introduce the recent progress on how machine learning has assisted in genome reannotation, contributed to the development of genetic manipulation tools, elucidated metabolic networks, and optimized bioprocesses in cyanobacterial and microalgal studies within the context of synthetic biology. Additionally, we analyze the current challenges facing this field and provide perspectives on further research directions for applying machine learning to synthetic biology in cyanobacteria and microalgae.