Designing a bacterial cell to churn out a useful chemical currently requires hundreds of trial-and-error experiments. This project aims to replace that guesswork with a computer simulation that models the entire life of a cell. The core problem is that existing computer models of cells are too simple. They only map steady-state metabolic reactions, missing the dynamic behaviour of genes and molecules over time. This forces researchers to rely on slow, iterative lab work to optimise a cell’s genome for producing compounds like biofuels or fine chemicals. The researchers will combine whole-cell models—which simulate every gene product and molecule across the cell’s lifecycle—with machine learning and techniques that edit many genes at once. If successful, this approach could radically speed up the design of bacterial cell factories. Instead of months of trial-and-error, a computer could predict the optimal set of genetic edits, and those edits could be made in a single step. The impact would be felt in industrial biotechnology: faster, cheaper routes to bio-based chemicals, fuels, and materials. The tools developed would be shared openly, allowing other labs to design production strains without deep expertise in modelling or genome editing.
View original technical description
Computer-aided design (CAD) is widely used across engineering processes. Applying CAD and computational approaches in engineering biology is a major ambition, as methods based only on biological knowledge require numerous experimental trial-and-error iterations. This is the case for many engineering biology applications, including metabolic engineering, where the aim is to redirect cellular metabolism to produce desired compounds (for example fine chemicals and biofuels) by modifying biosynthetic pathways. Genome-scale metabolic models (GEMs) are computational representations of metabolic fluxes in a cell, which can be used to predict how to best edit a genome (i.e. which genes to remove or insert) to optimise the production of the compound of interest. GEMs are however limited in their description of cellular processes, as they only focus on metabolic interactions at steady state. Whole-cell models (WCMs) are state-of-the-art cell representations that account for the function and dynamics of every gene product and molecule over the cell life cycle. WCMs offer wide-ranging and not yet fully explored opportunities to link genotypes to phenotypes, unlocking biological discovery for a range of biotechnological applications. We propose here to realise the potential of WCMs, together with modern Machine Learning and multiplexed genome editing techniques, to accelerate the engineering of bacterial cells that can be used as optimal production strains for metabolic engineering. Success will deliver radical changes in the way cell factories are designed and programmed, developing tools which interdisciplinary communities can access to deliver competitive bio-based solutions.
Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.
Is something wrong? Let us know