LBVS: an online platform for ligand-based virtual screening using publicly accessible databases.
Mol Divers
; 18(4): 829-40, 2014 Nov.
Article
en En
| MEDLINE
| ID: mdl-25182364
ABSTRACT
Abundant data on compound bioactivity and publicly accessible chemical databases increase opportunities for ligand-based drug discovery. In order to make full use of the data, an online platform for ligand-based virtual screening (LBVS) using publicly accessible databases has been developed. LBVS adopts Bayesian learning approach to create virtual screening models because of its noise tolerance, speed, and efficiency in extracting knowledge from data. LBVS currently includes data derived from BindingDB and ChEMBL. Three validation approaches have been employed to evaluate the virtual screening models created from LBVS. The tenfold cross validation results of twenty different LBVS models demonstrate that LBVS achieves an average AUC value of 0.86. Our internal and external testing results indicate that LBVS is predictive for lead identifications. LBVS can be publicly accessed at http//rcdd.sysu.edu.cn/lbvs.
Texto completo:
1
Colección:
01-internacional
Banco de datos:
MEDLINE
Asunto principal:
Interfaz Usuario-Computador
/
Sistemas en Línea
Tipo de estudio:
Diagnostic_studies
/
Prognostic_studies
/
Screening_studies
Idioma:
En
Revista:
Mol Divers
Asunto de la revista:
BIOLOGIA MOLECULAR
Año:
2014
Tipo del documento:
Article
País de afiliación:
China