Pattern generation and recognition in synthetic biology
File(s)
Author(s)
Wachter, Georg
Type
Thesis
Abstract
Multicellular life requires both pattern formation and pattern recognition in order to form a functioning organism. In this thesis I set out to engineer both, in an attempt to harness these processes for synthetic biology, and to gain a deeper appreciation for biological complexity. I first attempt to engineer synthetic Turing patterns to generate diverse gene expression patterns in E. coli. I then set out to understand generated patterns using mathematical theory, especially under the consideration of growth. To advance gene expression pattern recognition, I build molecular neural networks in mammalian cells. To achieve this, I build molecular perceptrons, and chain them into two example neural networks; one forming a dual region classifier, the other a bandpass. Additionally, I lay the foundations for in vivo weight optimisation algorithms. While typically treated as separate issues, convergence of pattern generation and recognition could enable spatio-temporal control of morphogenesis, and eventually unlock macro-level design of biology.
Version
Open Access
Date Issued
2023-08-18
Date Awarded
01/11/2023
License URL
Advisor
Endres, Robert
Isalan, Mark
Weiss, Ron
Sponsor
Imperial College London
Grant Number
Schrödinger Scholarship
Publisher Department
Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)