QSRR Models to Predict Retention Indices of Cyclic Compounds of Essential Oils

Chemometric approach to correlations between retention parameters of non-polar HPLC columns and physicochemical characteristics for ampholytic substances of biological and pharmaceutical relevance

The correlations between the retention parameters of forty ampholytic, biologically active and/or pharmaceutically relevant substances (obtained for three non-polar HPLC columns at various compositions of mobile phases and pH conditions: 2.5, 7.0, 11.5) and their thirty-two physicochemical (calculated/spectral) characteristics were investigated by applying chemometric methods of analysis. In three cases (among seven cases considered), Quantitative Property-Retention Relation (QPRR) models meeting the predictive capability criteria were developed (the values of R², Q²_CV, Q²_Ext were higher than 0.76, 0.66 and 0.67, respectively, while values of RMSE_C, RMSE_CV and RMSE_Ext were lower than 0.51, 0.65 and 0.65 in each developed model). These models create a useful platform for predicting retention parameters of untested chemicals and, to some extent, gaining pharmaceutically valuable information on the biologically active ampholytic substances based on their properties and the conditions of chromatographic separation.

AN IMPROVED QSPR STUDY OF REVERSE FACTOR OF NANOPARTICLES IN ROADSIDE ATMOSPHERE ON KERNEL PARTIAL LEAST SQUARES AND GENETIC ALGORITHM

Thermal desorption-comprehensive two-dimensional gas chromatography high-resolution time-of-flight mass spectrometry (TD–GC × GC–HRTOF-MS) is one of the most powerful tools in analytical nanoparticle compounds. Genetic algorithm and partial least square (GA-PLS) and kernel PLS (GA-KPLS) models were used to investigate the correlation between reverse factor (RF) and descriptors for 50 nanoparticles fraction with a diameter of 29–58 nm in roadside atmosphere which obtained by TD–GC×GC–HRTOF-MS. The correlation coefficient leave-group-out cross validation (LGO-CV (Q2)) of prediction for the GA-PLS and GA-KPLS models for training and test sets were (0.761 and 0.718) and (0.825 and 0.814), respectively, revealing the reliability of these models. This is the first research on the quantitative structure-property relationship (QSPR) of the nanoparticles in roadside atmosphere using the GA-PLS and GA-KPLS.

Chemometric model for predicting retention indices of constituents of essential oils

Quantitative structure-retention relationships (QSRRs) model was developed for predicting the gas chromatography retention indices of 169 constituents of essential oils. The ordered predictors selection algorithm was used to select three descriptors (one constitutional index and two edge adjacency indices) from 4885 descriptors. The final QSRR model (model M3) with three descriptors was internal and external validated. The leave-one-out cross-validation, leave-many-out cross-validation, bootstrapping, and y-randomization test indicated the final model is robust and have no chance correlation. The external validations indicated that the model M3 showed a good predictive power. The mechanistic interpretation of QSRR model was carried out according to the definition of descriptors. The results show that the larger molecular weight, the greater the values of retention indices. More compact structures have stronger intermolecular interactions between the components of essential oils and the capillary column. Therefore, the result meets the five principles recommended by the Organization for Economic Co-operation and Development (OECD) for validation of QSRR model, and it is expected the model can effectively predict retention indices of the essential oils.

Quantitative structure-retention relationship for retention behavior of organic pollutants in textile wastewaters and landfill leachate in LC-APCI-MS

INTRODUCTION

A quantitative structure-retention relation (QSRR) study was conducted on the retention times of organic pollutants in textile wastewaters and landfill leachate which was obtained by liquid chromatography-reversed phase-atmospheric pressure chemical ionization-mass spectrometry.

METHODS

The genetic algorithm was used as descriptor selection and model development method. Modeling of the relationship between selected molecular descriptors and retention time was achieved by linear (partial least square) and nonlinear (Levenberg-Marquardt artificial neural network, L-M ANN) methods. Linear and nonlinear models provide good results whereas more accurate results were obtained by the L-M ANN model.

CONCLUSION

This is the first research on the QSRR of the organic pollutants in textile wastewaters and landfill leachate against the retention time.

Application of artificial neural network to predict the retention time of drug metabolites in two-dimensional liquid chromatography

Genetic algorithm and partial least square (GA-PLS) and Levenberg-Marquardt artificial neural network (L-M ANN) techniques were used to investigate the correlation between retention time and descriptors for drug metabolites which obtained by two-dimensional liquid chromatography. The applied internal (leave-group-out cross validation (LGO-CV)) and external (test set) validation methods were used for the predictive power of four models. Both methods resulted in accurate prediction whereas more accurate results were obtained by L-M ANN model. The best model obtained from L-M ANN showed a good R(2) value (determination coefficient between observed and predicted values) for all compounds, which was superior to GA-PLS models.

pK(a) modelling and prediction of drug molecules through GA-KPLS and L-M ANN

Genetic algorithm and partial least square (GA-PLS), kernel PLS (GA-KPLS) and Levenberg- Marquardt artificial neural network (L-M ANN) techniques were used to investigate the correlation between dissociation constant (pK(a) ) and descriptors for 60 drug compounds. The applied internal (leave-group-out cross validation (LGO-CV)) and external (test set) validation methods were used for the predictive power of models. Descriptors of GA-KPLS model were selected as inputs in L-M ANN model. The results indicate that L-M ANN can be used as an alternative modeling tool for quantitative structure-property relationship (QSPR) studies.