ART: A machine learning Automated Recommendation Tool for synthetic biology

TL;DR

ART is a machine learning-based tool that systematically guides synthetic biology design by recommending strains and predicting their production levels, reducing development time without requiring full mechanistic understanding.

Contribution

The paper introduces ART, a novel machine learning and probabilistic modeling tool that automates and optimizes strain design in synthetic biology.

Findings

Successfully applied to simulated data and real metabolic engineering projects.

Provides probabilistic predictions of production levels for recommended strains.

Reduces development time in synthetic biology workflows.

Abstract

Biology has changed radically in the last two decades, transitioning from a descriptive science into a design science. 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, and fatty acids. Finally, we discuss the limitations of this approach, and the practical consequences of the underlying assumptions failing.