In:
Canadian Journal of Chemistry, Canadian Science Publishing, Vol. 90, No. 8 ( 2012-08), p. 640-651
Abstract:
Quantitative structure–property relationship (QSPR) studies were performed for the prediction of gas-phase reduced ion mobility constants (K 0 ) of diverse compounds based on three-dimensional (3D) molecular structure representation. The entire set of 159 compounds was divided into a training set of 120 compounds and a test set of 39 compounds according to Kennard and Stones algorithm. Multiple linear regression (MLR) analysis was employed to select the best subset of descriptors and to build linear models, whereas nonlinear models were developed by means of an artificial neural network (ANN). The obtained models with five descriptors involved show good predictive power for the test set: a squared correlation coefficient (R 2 ) of 0.9029 and a standard error of estimation (s) of 0.0549 were achieved by the MLR model, whereas by the ANN model, R 2 and s were 0.9292 and 0.496, respectively. The results of this study compare favorably to previously reported prediction methods for the ion mobility constants. In addition, the descriptors used in the models are discussed with respect to the structural features governing the mobility of the compounds.
Type of Medium:
Online Resource
ISSN:
0008-4042
,
1480-3291
Language:
English
Publisher:
Canadian Science Publishing
Publication Date:
2012
detail.hit.zdb_id:
1482256-8
Permalink