New solubility models based on descriptors derived from the detour matrix
New molecular descriptors were derived from already-known descriptors obtained from the Detour Matrix (also known as Maximal Topological Distance Matrix or Maximum Path Matrix) and applied to the prediction of aqueous solubility of 46 structurally heterogeneous compounds, constructing one-variable m...
Guardado en:
| Autores principales: | , , |
|---|---|
| Formato: | Articulo |
| Lenguaje: | Inglés |
| Publicado: |
2006
|
| Materias: | |
| Acceso en línea: | http://sedici.unlp.edu.ar/handle/10915/164781 |
| Aporte de: |
| Sumario: | New molecular descriptors were derived from already-known descriptors obtained from the Detour Matrix (also known as Maximal Topological Distance Matrix or Maximum Path Matrix) and applied to the prediction of aqueous solubility of 46 structurally heterogeneous compounds, constructing one-variable models through linear regression. The correlation coefficients between these descriptors and the experimental values of solubility were compared to those obtained with more than 1,600 widely-used descriptors included in commercial software Dragon, confirming the very good performance of the proposed descriptors. The best-performance descriptors were then applied, in combination with Dragon’s descriptors, to generate two five-variable models for the estimation of solubility. The F-Statistical and the p-value for this models confirmed high statistical significance. We also present the distribution of molecular weights, solubility values, number of H donors, number of H acceptors and number of heteroatoms for the 46 compounds employed, which show molecular diversity. The results indicate that the proposed descriptors can be applied in QSAR and QSPR studies. |
|---|