Synthetic biology allows us to bioengineer cells to synthesize novel valuable molecules such as renewable biofuels or anticancer drugs. However, traditional synthetic biology approaches involve ad-hoc engineering practices, which lead to long development times. Here, we present the Automated Recommendation Tool (ART), a tool that leverages machine learning and probabilistic modeling techniques to guide synthetic biology in a systematic fashion, without the need for a full mechanistic understanding of the biological system. Using sampling-based optimization, ART provides a set of recommended strains to be built in the next engineering cycle, alongside probabilistic predictions of their production levels. We demonstrate the capabilities of ART on simulated data sets, as well as experimental data from real metabolic engineering projects producing renewable biofuels, hoppy flavored beer without hops, fatty acids, and tryptophan. Finally, we discuss the limitations of this approach, and the practical consequences of the underlying assumptions failing.

合成生物学使我们能够对细胞进行生物工程，以合成新的有价值的分子，例如可再生生物燃料或抗癌药物。然而，传统的合成生物学方法涉及临时工程实践，这导致了很长的开发时间。在这里，我们介绍了自动推荐工具 (ART)，这是一种利用机器学习和概率建模技术以系统方式指导合成生物学的工具，而无需对生物系统进行全面的机械理解。使用基于采样的优化，ART 提供了一组在下一个工程周期中构建的推荐菌株，以及对其生产水平的概率预测。我们在模拟数据集上展示了 ART 的能力，以及来自生产可再生生物燃料、不含啤酒花、脂肪酸和色氨酸的啤酒花风味啤酒的真实代谢工程项目的实验数据。最后，我们讨论了这种方法的局限性，以及潜在假设失败的实际后果。