Related Experiment Videos
A radial-distribution-function approach for predicting rodent carcinogenicity
Aliuska Helguera Morales1, Miguel Angel Cabrera Pérez, Maykel Pérez González
1Department of Chemistry, Central University of Las Villas, Santa Clara, Villa Clara, 54830, Cuba.
Journal of Molecular Modeling
|January 20, 2006
Summary
This study introduces a new method using Radial-Distribution-Function (RDF) descriptors to predict chemical carcinogenicity. The RDF approach demonstrated superior accuracy in identifying carcinogenic and non-carcinogenic compounds compared to other computational methods.
Area of Science:
- Computational chemistry
- Toxicology
- Cheminformatics
Background:
- Predicting chemical carcinogenicity is crucial for drug safety and environmental health.
- Existing computational methods have limitations in accurately classifying carcinogenic potential.
Purpose of the Study:
- To develop and validate a predictive model for carcinogenic activity using Radial-Distribution-Function (RDF) descriptors.
- To compare the efficacy of RDF descriptors against eight other established computational methodologies.
Main Methods:
- A discriminant model was constructed using 188 compounds.
- Radial-Distribution-Function (RDF) descriptors were employed for feature extraction.
- Model performance was assessed via resubstitution cross-validation and an independent test set.
Main Results:
- The model achieved 76.4% overall classification for carcinogenic and 72.5% for non-carcinogenic chemicals.
- Cross-validation and test set validation yielded good classification rates of 79.3% and 72.5%, respectively.
- RDF descriptors outperformed Constitutional, Molecular walks counts, Galvez topological charge indices, 2D autocorrelations, Randić molecular profiles, Geometrical, 3D-MORSE, and WHIM methods.
Conclusions:
- Radial-Distribution-Function (RDF) descriptors provide a robust and effective approach for predicting chemical carcinogenicity.
- The developed RDF-based model shows promising predictive power and offers an advantage over several other computational techniques.