Use this URL to cite or link to this record in EThOS:
Title: Understanding stability of protein-protein complexes
Author: Agius, R.
ISNI:       0000 0004 5365 047X
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2015
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
For all living organisms, macromolecular interactions facilitate most of their natural functions. Alterations to macromolecular structures through mutations, can affect the stability of their interactions, which may lead to unfavourable phenotypes and disease. Presented here, are a number of computational methods aimed at uncovering the principles behind complex stability - as described by binding affinity and dissociation rate constants. Several factors are known to govern the stability of protein-protein interactions, however, no one factor dominates, and it is the synergistic effect of a number of contributions, which amount to the affinity, and stability of a complex. The characterization of complex stability can thus be presented as a two-fold problem; modelling the individual factors and modelling the synergistic effect of the combination of such individual factors. Using machine learning as a central framework, empirical functions are designed for estimating affinity, dissociation rates and the effects of mutations on these properties. The performance of all models is in turn benchmarked on experimental data available from the literature and carefully curated datasets. Firstly, a wild-type binding free energy prediction model is designed, composed of a diverse set of stability descriptors, which account for flexibility and conformational changes undergone by the complex in question. Similarly, models for estimating the effects of mutations on binding affinity are also designed and benchmarked in a community-wide blind trial. Emphasis here is on the detection of a small subset of mutations that are able to enhance the stability of two de novo protein drugs targeting the flu virus hemagglutinin. Probing further the determinants of stability, a set of descriptors that link hotspot residues with the off-rate of a complex are designed, and applied to models predicting changes in off-rate upon mutation. Finally, the relationship between the distribution of hotspots at protein interfaces, and the rate of dissociation of such interfaces, is investigated.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available