Exploring and expanding the solenoid protein universe
File(s)
Author(s)
Nikov, Georgi Ivanov
Type
Thesis
Abstract
Solenoid proteins are a class of tandem repeat proteins with a unique elongated and modular architecture, which lends itself to design and engineering. Design and engineering efforts benefit from well-annotated solenoid repeat data. Although there are databases of solenoid proteins and methods for their detection, there is room for improvement in the coverage of solenoid protein space and the quality of solenoid repeat annotations. In this thesis, a new convolutional neural network method for solenoid protein detection and annotation was developed: SOLeNNoID, and competitive performance against current state-of-the-art-methods was demonstrated. SOLeNNoID was then used to discover hundreds to thousands of solenoid proteins in the Protein Data Bank and the AlphaFold Database and the trends in these findings were discussed. Finally, to extend beta-solenoid design work from the Murray group, multiple beta-solenoid designs were evaluated in the laboratory: consensus hexapeptide repeat proteins, and SynRFR variants with two adjacent loops, SpyTag/SpyCatcher moieties, and truncations designed for spontaneous self-assembly. Taken together the solenoid protein tools, databases and experimental data in this thesis help to explore and expand the solenoid protein universe and could have applications in solenoid protein biotechnology and bioinformatics. All of the code produced in this thesis can be found at https://github.com/gnik2018.
Version
Open Access
Date Issued
2022-10-27
Date Awarded
01/06/2023
License URL
Advisor
Murray, James
Sponsor
Engineering and Physical Sciences Research Council
Publisher Department
Life Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)