Engineering Cellular Metabolism using Machine Learning
Leveraging advanced mechanistic modelling and the generation of high-quality multi-dimensional data sets, machine learning is becoming an integral part of understanding and engineering living systems.
Among the different types of mechanistic models for simulating metabolism, genome-scale models are one of the most popular approaches. Yet, the predictive power of genome-scale models is often hampered by the limited knowledge and data available for the parameters affecting metabolic regulation. Here we present how mechanistic and machine learning models can complement each other in a combined approach enabling predictive engineering of yeast metabolism.
From a single data-generation cycle, we demonstrate that this approach enables successful forward engineering of complex native and heterologous metabolisms in yeast, with the best machine learning-guided design recommendations improving titers and productivities compared to the best designs used for algorithm training. Taken together, this presentation will highlight the power of combining mechanistic and machine learning models to effectively direct metabolic engineering efforts.
1. Novo Nordisk Foundation Center for Biosustainability (Technical University of Denmark, Kgs. Lyngby, Denmark):
- Jie Zhang
- Søren Petersen
- Christine Pedersen
- Michael Krogh Jensen
2. Joint BioEnergy Institute, Emeryville, CA, USA:
- Hector Garcia-Martin
3. TeselaGen Biotechnology, San Francisco, CA 94107, USA:
- Mike Fero
Archived event
This event has already happened.
Documents & links
-
Engineering Cellular Metabolism using Machine Learning
Leveraging advanced mechanistic modelling and the generation of high-quality multi-dimensional data sets, machine learning is becoming an integral part of understanding and engineering living systems.
Among the different types of mechanistic models for simulating metabolism, genome-scale models are one of the most popular approaches. Yet, the predictive power of genome-scale models is often hampered by the limited knowledge and data available for the parameters affecting metabolic regulation. Here we present how mechanistic and machine learning models can complement each other in a combined approach enabling predictive engineering of yeast metabolism. From a single data-generation cycle, we demonstrate that this approach enables successful forward engineering of complex native and heterologous metabolisms in yeast, with the best machine learning-guided design recommendations improving titers and productivities compared to the best designs used for algorithm training. Taken together, this presentation highlighted the power of combining mechanistic and machine learning models to effectively direct metabolic engineering efforts.
Download