People involved: Olivier Borkowski
Keywords: cell free, metabolic engineering, machine learning, bacteria, automation
This project aims to challenge the problem of optimisation of a biological pathway for the production of high value chemical production in vivo and in vitro. We will construct, using a cell free platform and in vivo measurements, an optimised bacterial strain and in vitro kits for a robust production of flavonols. The resulting strains/kits will offer a high potential for industrial application. For the first time, cell free technology will be used to cover the design space of a metabolic pathway and will be coupled with machine learning in a design/build/test/learn cycle to optimize this pathway before transformation in E. coli. The flavonol pathway will be optimized though maximization of enzymes productions and efficiencies with a minimal burden for the cell. The results obtained for in vivo production will be used to also provide an optimized pathway for flavonols production in vitro. For both in vivo and in vitro optimizations, the initial step will consist of measuring a large library of enzymes sequences / level of production in different strains lysates (cell free conditions) to explore the design space of the flavonols pathway. Given the scope of the design space, the initial measurements will be used to set the initiation conditions for a maximisation process using cycle of design/build/test/learn. The maximisation process will use directed mutations coupled with machine learning to understand and optimised the sequences and the production level of the enzymes of interest. In both steps, sequencing reads will be used to provide the data necessary for the machine learning process.