Related Experiment Video
Updated: Dec 7, 2025
11:22
Automated Robotic Liquid Handling Assembly of Modular DNA Devices
Published on: December 1, 2017
12.7K
A machine learning Automated Recommendation Tool for synthetic biology
Tijana Radivojević1,2,3, Zak Costello1,2,3, Kenneth Workman1,3,4
1DOE Agile BioFoundry, Emeryville, CA, 94608, USA.
Nature Communications
|September 26, 2020
Summary
Automated Recommendation Tool (ART) streamlines synthetic biology by using machine learning to suggest cell engineering strategies. This accelerates the development of valuable molecules like biofuels and drugs.
Area of Science:
- Synthetic biology
- Metabolic engineering
- Computational biology
Background:
- Traditional synthetic biology relies on time-consuming, ad-hoc engineering.
- A lack of mechanistic understanding often hinders biological system optimization.
- Developing novel molecules like biofuels and pharmaceuticals requires efficient engineering.
Purpose of the Study:
- To introduce the Automated Recommendation Tool (ART) for systematic synthetic biology.
- To guide strain engineering using machine learning and probabilistic modeling.
- To accelerate the development of bioengineered molecules without full mechanistic insight.
Main Methods:
- ART employs sampling-based optimization.
- It utilizes machine learning and probabilistic modeling.
- The tool provides recommended strains and predicts production levels.
Main Results:
- ART's efficacy was demonstrated on simulated and experimental datasets.
- Successful applications include producing biofuels, novel beer flavors, fatty acids, and tryptophan.
- The tool guides strain selection for improved molecule synthesis.
Conclusions:
- ART offers a systematic approach to synthetic biology, reducing development time.
- It enables data-driven strain engineering without complete biological system knowledge.
- The tool has broad applicability in metabolic engineering and biomanufacturing.
