Use this URL to cite or link to this record in EThOS:
Title: Computing free energy, binding and competition within Fragment Based Drug Discovery
Author: Cabedo Martinez, Ana
ISNI:       0000 0004 5991 8991
Awarding Body: University of Southampton
Current Institution: University of Southampton
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The development of JAFS, a new computational method to study the binding geometries of small fragment molecules to protein cavities, estimate their binding affnities and analyse how they compete for a common protein binding site, all in the context of Fragment Based Drug Discovery, is presented in this thesis. Fragment Based Drug Discovery is an approach to drug development which studies the binding of small ligands (fragments) forming high quality interactions with their target. Further optimization of these fragments into drug-like molecules, adding functionalities to increase affnity while controlling other relevant properties such as toxicity and absorption then takes place. JAFS studies the binding of fragments to their target proteins. The JAFS method consists of the execution and analysis of Monte Carlo simu-lations of fragments (and waters) in the binding cavities of proteins with an added degree of freedom which accounts for the scaling of the interaction energy of the fragment (and water). Sampling of states at very low interaction energies gives a boost in fragment con?gurational sampling while competition between di?erent fragments to remain at unscaled (high) interaction energies at a given binding site provides information on their relative binding a?nities. JAFS is built on the JAWS formulation for water binding to protein cavities. The performance of the JAFS method on a range of different test cases (T4 Lyzozyme, Major Urinary Protein I, Cyclin Dependent Kinase 2 and Heat Shock Protein 90) was studied. JAFS is divided in two protocols to rank fragments by affnity and locate binding geometries, respectively. The ranking of fragments by affnity to a common protein target was satisfactory (as compared to experimen-tal data) for the simpler systems (T4 Lyzozyme and Major Urinary Protein I). However, more demanding systems proved problematic, where the ranking of nine different ligands to the binding site of Cyclin Dependent Kinase 2 provided results unrelated to experimental binding affnities. Studying pose generation in sets of five repeats per simulation, the crystal binding geometry of every fragment studied was found in at least one of the re-peats, without providing any previous information on the system (such as the presence or location of water mediated interactions or the hydration state of the cavity). Consistency between repeats was however found to be problematic and no method is currently able to select the optimal binding geometry among all the gen-erated poses. Suggestions are given for further developments which would provide a methodology to rank poses.
Supervisor: Essex, Jonathan Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available