The development of novel scoring methods for virtual screening
The development of novel scoring methods for virtual screening
Protein ligand docking is a valuable technique in the field of Structure Based Drugs Design. One of its applications is to select one or few candidates from a set of thousands or million compounds, often known as Virtual Screening. The choice of the current scoring function is of paramount importance, and currently available scoring functions are most often implemented to optimise reproduction of known binding modes, or experimental free energies. The vast amount of information contained in the fact that most ligands do not, indeed, bind to a given protein is discarded, building a considerable bias in these functions. The parameterisation method proposed in this work attempts to overcome said issue, by using a set of decoys among which the known ligand is hidden. The parameters of the scoring function are chosen by a Genetic Algorithm so to maximize the known ligand’s ranking. The generality of the function is maintained by the constraint of simultaneously optimising several of these sets: the influence of the training set’s size on the quality and robustness of the function is assessed. Also the effect of using different reference poses for the known ligand, either experimental or re=generated by docking, is investigated. Finally, the immediate applicability of the method is shown by reparameterising commercially available scoring functions (GOLDSCORE, CHEMSCORE); a final attempt is made at creating scoring functions tailored to specific protein families; interpreting the results in terms of what is known of their ligand-protein interactions. The results show that such a function is able to discriminate between known actives and decoys, and would therefore be a valuable addition to the computational chemist bag of tricks to select those molecules able to turn into successful drugs.
University of Southampton
Fenu, Luca Antonio
1a49c83b-d586-4513-9b94-4801875c8c07
2007
Fenu, Luca Antonio
1a49c83b-d586-4513-9b94-4801875c8c07
Fenu, Luca Antonio
(2007)
The development of novel scoring methods for virtual screening.
University of Southampton, Doctoral Thesis.
Record type:
Thesis
(Doctoral)
Abstract
Protein ligand docking is a valuable technique in the field of Structure Based Drugs Design. One of its applications is to select one or few candidates from a set of thousands or million compounds, often known as Virtual Screening. The choice of the current scoring function is of paramount importance, and currently available scoring functions are most often implemented to optimise reproduction of known binding modes, or experimental free energies. The vast amount of information contained in the fact that most ligands do not, indeed, bind to a given protein is discarded, building a considerable bias in these functions. The parameterisation method proposed in this work attempts to overcome said issue, by using a set of decoys among which the known ligand is hidden. The parameters of the scoring function are chosen by a Genetic Algorithm so to maximize the known ligand’s ranking. The generality of the function is maintained by the constraint of simultaneously optimising several of these sets: the influence of the training set’s size on the quality and robustness of the function is assessed. Also the effect of using different reference poses for the known ligand, either experimental or re=generated by docking, is investigated. Finally, the immediate applicability of the method is shown by reparameterising commercially available scoring functions (GOLDSCORE, CHEMSCORE); a final attempt is made at creating scoring functions tailored to specific protein families; interpreting the results in terms of what is known of their ligand-protein interactions. The results show that such a function is able to discriminate between known actives and decoys, and would therefore be a valuable addition to the computational chemist bag of tricks to select those molecules able to turn into successful drugs.
Text
1138454.pdf
- Version of Record
More information
Published date: 2007
Identifiers
Local EPrints ID: 466478
URI: http://eprints.soton.ac.uk/id/eprint/466478
PURE UUID: ae15f8c6-4fc2-4882-81be-0f5d9d7b8b6e
Catalogue record
Date deposited: 05 Jul 2022 05:18
Last modified: 16 Mar 2024 20:43
Export record
Contributors
Author:
Luca Antonio Fenu
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics