Scoring functions for fragment-based drug discovery.
Methods Mol Biol
; 1289: 101-15, 2015.
Article
em En
| MEDLINE
| ID: mdl-25709036
ABSTRACT
Fragment-based drug design represents a challenge for computational drug design because almost inevitably fragments will be weak binders to the biomolecular targets of a specific disease, and the performances of the scoring functions for weak binders are usually poorer than those for the stronger binders. This protocol describes how to predict the binding modes and binding affinities of fragments towards their binding partner with our refined AutoDock scoring function incorporating a quantum chemical charge model, namely, the restrained electrostatic potential (RESP) model. This scoring function was calibrated by robust regression analysis and has been demonstrated to perform well for general classes of protein-ligand interactions and for weak binders (with root-mean square of error of about 2.1 kcal/mol).
Texto completo:
1
Base de dados:
MEDLINE
Assunto principal:
Proteínas
/
Modelos Moleculares
/
Bibliotecas de Moléculas Pequenas
/
Descoberta de Drogas
/
Ligantes
/
Modelos Químicos
Idioma:
En
Ano de publicação:
2015
Tipo de documento:
Article