Whole cell modelling for bacteria E.coli

点击翻译按钮获取中文摘要

One of the greatest challenges that the fields of systems and synthetic biology are facing at the moment is the ability to make the engineering and production of designed genomes across applications more efficient. In order to achieve this, there is a particular need for better mathematical models of cells. These models have taken the name of whole-cell computational models (WCMs), and the aim is to design them such that the function of all genes and molecules are taken into account. Addressing this challenge would contribute greatly to a better understanding of different diseases and the creation of personalised treatments. WCMs will allow us to predict phenotypes directly from genotypes and therefore investigate changes in the cell after being exposed to different environmental factors.This project will focus on improving the recently developed WCM for bacteria E. coli [1]. This model was designed by the members of the Covert lab from the University of Stanford and despite representing a breakthrough in cell engineering and design, there are still some challenges that need to be addressed. Currently, this model is taking into account only half of the genes of the cell. Without a WCM that includes all functionalities of the cell, it is hard to accurately predict the growth in the presence of antibiotics or other external stimulus, and to simulate gene knockouts for genes that are not implemented. This project attempts to solve this challenge by implementing novel ways in which WCMs are created and improved, based on artificial intelligence (AI) technologies. These models have the advantage of being easier to prototype compared to a mathematical model, especially one that is combining different techniques such as ordinary differential equations, stochastic modelling and geometric analysis (like the current WCMs).In addition, another important aim of this project is to unravel the link between the genotype and the phenotype of the cells. A combination of techniques such as unsupervised and deep learning will allow us to answer questions such as 'How do multiple gene knock-out/in affect the growth and metabolism of the cell?', for the genes that have already been modelled. Another big challenge that the community is facing when designing WCM is the ability to validate the mathematical models against the experimental data. This project will start to address this issue by deriving data-centric sub-models (using for example particle swarm optimisation techniques), for the biological processes that provide easy access to experimental data, such as RNA expression.The intersection of systems/synthetic biology and AI is a new area of research that has a lot of potential. The WCM for bacteria E.coli project is particularly interesting for starting to explore the applications of AI in systems biology because of the large amount of the data that has been gathered from the community during the past decades.This project falls within the EPSRC Synthetic Biology research area.The collaborators of this project are: Dr. Lucia Marucci (main supervisor), Prof. Claire Grierson (main supervisor), Dr. Thomas Gorochowski (co-supervisor, Royal Society University Research Fellow in the School of Biological Sciences), Dr. Wei Pang (co-supervisor, Associate Professor in Computer Science at Heriot-Watt University and an Honorary Senior Lecturer at Aberdeen University) and the Covert lab group from the University of Stanford.[1] Macklin, Derek & Ahn-Horst, Travis & Choi, Heejo & Ruggero, Nicholas & Carrera, Javier & Mason, John & Sun, Gwanggyu & Agmon, Eran & Defelice, Mialy & Maayan, Inbal & Lane, Keara & Spangler, Ryan & Gillies, Taryn & Paull, Morgan & Akhter, Sajia & Bray, Samuel & Weaver, Daniel & Keseler, Ingrid & Karp, Peter & Covert, Markus. (2020). Simultaneous cross-evaluation of heterogeneous E. coli datasets via mechanistic simulation. Science (New York, N.Y.). 369. 10.1126/science.aav3751