QSRR Models Research Articles

Quantitative Structure Retention-Relationship Modeling: Towards an Innovative General-Purpose Strategy.

Reversed-Phase Liquid Chromatography (RPLC) is a common liquid chromatographic mode used for the control of pharmaceutical compounds during their drug life cycle. Nevertheless, determining the optimal chromatographic conditions that enable this separation is time consuming and requires a lot of lab work. Quantitative Structure Retention Relationship models (QSRR) are helpful for doing this job with minimal time and cost expenditures by predicting retention times of known compounds without performing experiments. In the current work, several QSRR models were built and compared for their adequacy in predicting the retention times. The regression models were based on a combination of linear and non-linear algorithms such as Multiple Linear Regression, Support Vector Regression, Least Absolute Shrinkage and Selection Operator, Random Forest, and Gradient Boosted Regression. Models were built for five pH conditions, i.e., at pH 2.7, 3.5, 6.5, and 8.0. In the end, the model predictions were combined using stacking and the performances of all models were compared. The k-nearest neighbor-based application domain filter was established to assess the reliability of the prediction for further compound prioritization. Altogether, this study can be insightful for analytical chemists working with RPLC to begin with the computational prediction modeling such as QSRR to predict the separation of small molecules.

QSRR models for predicting the retention indices of VOCs in different datasets using an efficient variable selection method coupled with artificial neural network modeling: ANN-based QSPR modeling

QSRR models for predicting the retention indices of VOCs in different datasets using an efficient variable selection method coupled with artificial neural network modeling: ANN-based QSPR modeling

The evaluation of the anticancer activity of the Biginelli hybrids and pharmacokinetic profiling based on their retention parameters

The present investigation gives an insight into the evaluation and analysis of the anticancer activity of the library of the Biginelli hybrids using the appropriate QSRR approach. Using the RP TLC method retention parameters of tested compounds were obtained and examined to measure of lipophilicity of investigated molecules. The compounds were examined in seven different cancer cell lines and their IC50 and% of inhibition of cell proliferation at 100 µM were established. These experimental values as well as appropriate molecular descriptors are included in QSRR analysis. For this purpose, the variable selection was made, PCA and HCA were carried out, nine MR models were developed and ranked. The quality of the established models was confirmed through internal and external statistical validation. The goal was to define the main differences and similarities between three groups of the tested Biginelli hybrids to -find out which molecular features affect lipophilicity the most and which are crucial for the development of high-quality QSRR models. Tetrahydropyrimidines with butyl (11) and benzyl fragment (19) possess the best anticancer activity and selectivity. Nowadays modern design of an active pharmaceutical ingredient includes specific requirements of rationalization to adapt physicochemical characteristics, pharmacological activity, and safety through structural changes of the molecule. We believe that the developed profile is a step forward in Biginelli chemistry and could be useful in the future synthesis of novel Biginelli-based compounds with significantly improved activities.

Study of chromatographic behavior of antibiotic drugs and their metabolites based on quantitative structure-retention relationships with the use of HPLC-DAD

The separation of eleven antibiotics and ten metabolites were studied using high performance liquid chromatography. The C18-PFP octadecyl with integral PFP, C18-AR octadecyl with integral phenyl, C18-HL octadecyl and phenyl phase were used as stationary phases. Mixtures of acetontrile–0.1 % formic acid in water were investigated as mobile phases. The elution order of the target compounds was similar for all four HPLC columns applied. The best separation was obtained using the column with the pentafluorophenylpropyl chain. In addition, in order to optimize the parameters of retention elution for the column and to predict the conditions for the best separation of the active compounds studied biologically the ChromSword software was used. To obtain reliable information of the physicochemical properties and to estimate the relative biological activity of a group of the studied analytes, the QSRR approach was applied. Molecular descriptors were calculated using the HyperChem software. The study was based on multiple linear regression and the results were presented as quantitative structure–retention relationships equations. The QSRR models were determined using 16 molecular descriptors mainly related to the dipole moment (μ), the solvent accessible surface area (SAS), the van der Waals surface area (VWS), the minimum charge (δmin) as well as the polar surface area (PSA). Moreover, structural descriptors of the target compounds were used to describe their chromatographic retention behavior under the optimized HPLC conditions.

Quantitative structure – retention relationships of amino acids on the amino acid- and peptide-silica stationary phases for liquid chromatography

Quantitative structure – retention relationships analysis was applied to investigate the molecular retention mechanism of proteinogenic and non-proteinogenic amino acids on the amino acid- and peptide-silica stationary phases. Twelve stationary phases with chemically bonded amino acids of different types (glycine, alanine, phenylalanine, leucine, methionine, aspartic acid, and N-(9-Fluorenylmethoxycarbonyl)-O-tert-butyl-l-tyrosine) and chains lengths (amino acid, dipeptide, and tripeptide) were tested. In order to compare chromatographic properties of the prepared materials with the conventional columns, the amino-bonded phases (laboratory-prepared and commercial one) were also studied. For each of analyte, the molecular descriptors were calculated using quantum mechanics method. The QSRR models were determined using 13 molecular descriptors mainly related to the surface area, hydrophobicity, polarity, ion-exchange and hydrogen bonding capabilities of the analytes. Finally, the prediction potency of the molecular modeling descriptors-based models was also independently studied for the tested stationary phases using 15 training set and 6 test set of amino acids.

Application of Ionic Liquids for the Determination of Lipophilicity Parameters Using TLC Method, and QSRR Analysis for the Antipsychotic Drugs.

Reversed-phase liquid chromatography may cause difficulties, especially in the case of basic drugs due to the strong silanophilic interactions in the partition mechanism. Recently, imidazolium-based ionic liquids additives appeared interesting and a convenient solution for suppressing the harmful effect of free residuals of silanol groups, allowing remodeling of the stationary/mobile-phase system, and thus improving the lipophilicity assessment process. The aim of the study was to evaluate the retention behavior of basic antipsychotics using various RP-LC systems, and compare them with data obtained from the modified ionic-liquids RP-TLC systems, and perform the QSRR analysis. Retention and lipophilicity parameters of diverse antipsychotics have been examined in various RP-LC systems. Lipophilicity indices were compared with miscellaneous computed logP values. Furthermore, a large number of molecular descriptors have been computed and compared using various medicinal chemistry software, in order to contribute to the analysis of QSRR. Designated correlation coefficients showed that lipophilicity parameters from TLC systems without [EMIM][BF4] additive correlates very poor with the calculated logPs indices, whereas the indices from the traditional HPLC and TLC systems (with [EMIM][BF4]) were clearly better. Furthermore, QSRR analysis performed for these experimentally obtained lipophilicity parameters showed significant relationships between the retention constants (ROM, logkw) and the in silico calculated physicochemical molecular descriptors. ILs additive may be a significant factor affecting the lipophilicity of basic compounds, thus their use may be favorable in lipophilicity assessment studies. QSRR models with ILs showed that they may be useful in searching/or predicting HPLC/TLC retention parameters for the new/other antipsychotic drugs.

Investigation and prediction of retention characteristics of imidazoline and serotonin receptor ligands and their related compounds on mixed-mode stationary phase

Investigation of the retention behavior of a wide range of analytes, 43 nitrogen containing heterocyclic and guanidine derivatives such, as imidazoline and serotonin receptor ligands or their related compounds, was performed on mixed-mode stationary phase in the combined reversed-phase (RP) and hydrophilic interaction liquid chromatography (HILIC) modes.Suitability of the linear retention modelling in the HILIC and RP modes was tested including separate contributions from adsorption and partition. For the HILIC retention, the partition model was found to provide better description compared with the adsorption model. In a wider range of the aqueous eluent volume fractions, φ(aq), retention was described as a function of volume fractions and total polarity of mobile phase using the mixed-mode retention modelling. The obtained results revealed that the shift of the chromatographic mode can be calculated from the change of total polarity of mobile phase in a multi-modal relation, logarithm of retention factor vs. total polarity, with the minimum value representing the turning point between the HILIC and the RP mode.Molecular properties of the investigated compounds that influence the retention behavior and the turning point were selected using Multiple Linear Regression (MLR) and Support Vector Machine (SVM). Slightly better statistical results were found for the logkwRP(aq)/MLR, logkwHILIC(org)/MLR, logkbHILIC(aq)/MLR, and φmin (aq)/SVM (RBF) QSRR models than for the logkwRP(aq)/SVM, logkwHILIC(org)/SVM, logkbHILIC(aq)/SVM, and φmin(aq)/MLR modelling. With this insight, it is possible to precisely define and predict the retention characteristics based on physico-chemical properties of imidazoline and piperazine related compounds.

QSRR modeling for the chromatographic retention behavior of some β-lactam antibiotics using forward and firefly variable selection algorithms coupled with multiple linear regression

The justified continuous emerging of new β-lactam antibiotics provokes the need for developing suitable analytical methods that accelerate and facilitate their analysis. A face central composite experimental design was adopted using different levels of phosphate buffer pH, acetonitrile percentage at zero time and after 15 min in a gradient program to obtain the optimum chromatographic conditions for the elution of 31 β-lactam antibiotics. Retention factors were used as the target property to build two QSRR models utilizing the conventional forward selection and the advanced nature-inspired firefly algorithm for descriptor selection, coupled with multiple linear regression. The obtained models showed high performance in both internal and external validation indicating their robustness and predictive ability. Williams-Hotelling test and student’s t-test showed that there is no statistical significant difference between the models’ results. Y-randomization validation showed that the obtained models are due to significant correlation between the selected molecular descriptors and the analytes’ chromatographic retention. These results indicate that the generated FS-MLR and FFA-MLR models are showing comparable quality on both the training and validation levels. They also gave comparable information about the molecular features that influence the retention behavior of β-lactams under the current chromatographic conditions. We can conclude that in some cases simple conventional feature selection algorithm can be used to generate robust and predictive models comparable to that are generated using advanced ones.

Chromatographic behavior of selected antibiotic drugs supported by quantitative structure-retention relationships

As antibiotic drugs have a wide variety of applications, there is a growing demand for their selective determination, a task for which chromatographic methods seem to be appropriate. With this end in view, chromatographic measurements were performed with the use of six different HPLC columns (ACE 5C18, IAM.PC.DD.2, octadecyl, phenyl, alkylamide and cholesteryl-alkylamide) to determine the logkw of selected antibiotic drugs (amoxicillin, cefatoxime, ciprofloxacin, fluconazole, gentamicin, clindamycin, linezolid and metronidazole). The retention behavior of the analytes was investigated as a function of different binary hydro-organic mobile phases containing 10–45% (v/v) acetonitrile. The studied compounds were separated under isocratic conditions. The best results of separation of the tested biologically active compounds were obtained on the commercially available C18 column. Special attention was dedicated to the study of interactions among the stationary phase, mobile phase and the analytes. Another goal was to selecting the best column for separation of the tested biologically active compounds. Finally, QSRR models together with stationary phase characterization provided reliable information on the properties and characteristics of studied columns.

QSRR Study of Substituted Phenol Compounds: Proposed Models for Inter-Column Comparison

In gas chromatography, the comparison of experimental results obtained from different columns is a complex problem. In this sense, the QSRR methodology could determine, with high quality of description and prediction, the chromatographic retention index, allowing to use this methodology for comparison of data obtained from different column polarities and even obtained from different laboratories. A set of phenolic compounds was studied by several QSRR models; however, the prediction of such models’ descriptors is complicated, to understand, thus limiting its application. In the present work, three QSRR models from inter-column combinations are studied by quantum chemical descriptors, such as electronic energy (and its logarithm) and the corrected electronic energy (and its logarithm), zero point energy (ZPE), dipole moment in x-, y-, and z-axes, the total dipole moment, the HOMO and LUMO energies, hardness (η), electronegativity (χ), and electrophilicity (ω). These QSRR models allow the comparison of experimental chromatographic retention index with different column polarities for phenolic compounds.

UHPLC-QTOF MS screening of pharmaceuticals and their metabolites in treated wastewater samples from Athens

After consumption, pharmaceuticals are excreted as parent compounds and/or metabolites in urine and faeces. Some are not completely removed during wastewater treatments, forcing sewage treatment plants (STPs) to apply alternative technologies to guarantee quality of treated water. To monitor the removal efficiency of STPs, not only unchanged compounds and metabolites have to be taken into account, but also formation of possible transformation products (TPs). In this work, QTOF MS has been used for screening metabolites/TPs of pharmaceuticals in effluent wastewater from Athens. A customised database was built with the exact masses of metabolites reported in literature for the parent drugs found in an initial screening. Additionally, TPs identified in previous degradation experiments performed at our laboratory were included. Up to 34 metabolites/TPs were detected for omeprazole, venlafaxine, clindamycin, clarithromycin, clopidogrel or dipyrone, among others. Seven corresponded to TPs whose reference standards were available at our lab, seven were TPs previously identified in laboratory degradation experiments, eight were TPs tentatively identified by QTOF MS without reference standards, and twelve TPs were discovered after using the common fragmentation pathway approach. Tentative identification of TPs was supported by prediction of their chromatographic retention time based on the use of advanced chemometric QSRR models.

Quantitative structure-(chromatographic) retention relationship models for dissociating compounds

The aim of this work was to develop mathematical models relating the hydrophobicity and dissociation constant of an analyte with its structure, which would be useful in predicting analyte retention times in reversed-phase liquid chromatography. For that purpose a large and diverse group of 115 drugs was used to build three QSRR models combining retention-related parameters (logkw—chromatographic measure of hydrophobicity, S—slope factor from Snyder-Soczewinski equation, and pKa) with structural descriptors calculated by means of molecular modeling for both dissociated and nondissociated forms of analytes. Lasso, Stepwise and PLS regressions were used to build statistical models. Moreover a simple QSRR equations based on lipophilicity and dissociation constant parameters calculated in the ACD/Labs software were proposed and compared with quantum chemistry-based QSRR equations. The obtained relationships were further used to predict chromatographic retention times. The predictive performances of the obtained models were assessed using 10-fold cross-validation and external validation. The QSRR equations developed were simple and were characterized by satisfactory predictive performance. Application of quantum chemistry-based and ACD-based descriptors leads to similar accuracy of retention times’ prediction.

Quantitative structure retention/activity relationships of biologically relevant 4-amino-7-chloroquinoline based compounds.

The chromatographic behaviour of series of 4-amino-7-chloroquinoline (4,7-ACQ) based compounds was studied by reversed-phase thin-layer chromatography (RPTLC) with binary mobile phases containing water and the organic modifiers, DMSO or acetone. The lipophilicity of the studied compounds was determined by extrapolation of retention parameters RM to pure water content in mobile phase. In order to obtain some basic insight into the chromatographic behaviour and structural features of investigated compounds, PCA was performed on both chromatographic data (RM values) and calculated 2D and 3D structural descriptors. Both QSRR and QSAR models were built by means of the partial least squares (PLS) statistical method. It was found that descriptors which encode hydrophobic (dispersive) interactions have positive influence on retention, while influence of descriptors encoding polar interactions was negative. According to the obtained PLS model for inhibition of botulinum neurotoxin serotype A light chain, hydrophobic interactions influence positively on the mechanism of action of the investigated 4,7-ACQ, while polar interactions are less favoured. Contrary, the results of PLS modelling of activity against Plasmodium falciparum strains (W2, D6 and TM91C235) indicate that higher polarity of 4,7-ACQ contribute to their higher antimalarial activity.

Theoretical Models and QSRR in Retention Modeling of Eight Aminopyridines.

In this article, retention modeling of eight aminopyridines (synthesized and characterized at the Faculty of Pharmacy) in reversed-phase high performance liquid chromatography (RP-HPLC) was performed. No data related to their retention in the RP-HPLC system were found. Knowing that, it was recognized as very important to describe their retention behavior. The influences of pH of the mobile phase and the organic modifier content on the retention factors were investigated. Two theoretical models for the dependence of retention factor of organic modifier content were tested. Then, the most reliable and accurate prediction of log k was created, testing multiple linear regression model-quantitative structure-retention relationships (MLR-QSRR) and support vector regression machine-quantitative structure-retention relationships (SVM-QSRR). Initially, 400 descriptors were calculated, but four of them (POM, log D, M-SZX/RZX and m-RPCG) were included in the models. SVM-QSRR performed significantly better than the MLR model. Apart from aminopyridines, four structurally similar substances (indapamide, gliclazide, sulfamethoxazole and furosemide) were followed in the same chromatographic system. They were used as external validation set for the QSRR model (it performed well within its applicability domain, which was defined using a bounding box approach). After having described retention of eight aminopyridines with both theoretical and QSRR models, further investigations in this field can be conducted.

Chemometric Analysis of Some Biologically Active Groups of Drugs on the Basis Chromatographic and Molecular Modeling Data.

In this work, three different groups of drugs such as 12 analgesic drugs, 11 cardiovascular system drugs and 36 "other" compounds, respectively, were analyzed with cluster analysis (CA), principal component analysis (PCA) and factor analysis (FA) methods. All chemometric analysis were based on the chromatographic parameters (logk and logk(w)) determined by means of high-performance liquid chromatography (HPLC) and also by molecular modeling descriptors calculated using various computer programs (HyperChem, Dragon, and the VCCLAB). The clustering of compounds were obtained by CA (using various algorithm as e.g. Ward method or unweighted pair-group method using arithmetic averages as well as Euclidean or Manhattan distance), and allowed to build dendrograms linked drugs with similar physicochemical and pharmacological properties were discussed. Moreover, the analysis performed for analyzed groups of compounds with the use of FA or PCA methods indicated that almost all information reached in input chromatographic parameters as well as in molecular modeling descriptors can be explained by first two factors. Additionally, all analyzed drugs were clustered according to their chemical structure and pharmacological activity. Summarized, the performed classification analysis of studied drugs was focused on similarities and differences in methods being used for chemometric analysis as well as focused abilities to drugs classification (clustering) according to their molecular structures and pharmacological activity performed on the basis of chromatographic experimental and molecular modeling data. Thus, the most important application of statistically important molecular descriptors taken from QSRR models to classification analysis allow detailed biological (pharmacological) classification of analyzed drugs.

Least absolute shrinkage and selection operator and dimensionality reduction techniques in quantitative structure retention relationship modeling of retention in hydrophilic interaction liquid chromatography

The objective of this study was to model the retention of nucleosides and pterins in hydrophilic interaction liquid chromatography (HILIC) via QSRR-based approach.Two home-made (Amino-P-C18, Amino-P-C10) and one commercial (IAM.PC.DD2) HILIC stationary phases were considered. Logarithm of retention factor at 5% of acetonitrile (logkACN) along with descriptors obtained for 16 nucleosides and 11 pterins were used to develop QSRR models. We used and compared the predictive performance of three regression techniques: partial least square (PLS), the least absolute shrinkage and selection operator (LASSO), and the LASSO followed by stepwise multiple linear regression.The highest predictive squared correlation coefficient (QLOOCV2) in PLS analysis was found for Amino-P-C10 (QLOOCV2=0.687) and IAM.PC.DD2 (QLOOCV2=0.506) and the lowest for IAM.PC.DD2 (QLOOCV2=−0.01). Much higher values were obtained for the LASSO model. The QLOOCV2 equaled 0.9 for Amino-P-C10, 0.66 for IAM.PC.DD2 and 0.59 for Amino-P-C18. The combination of LASSO with stepwise regression provided models with comparable predictive performance as the LASSO, however with possibility of calculating the standard error of estimates.The use of LASSO itself and in combination with classical stepwise regression may offer greater stability of the developed models thanks to more smooth change of coefficients and reduced susceptibility towards chance correlation. Application of QSRR-based approach, along with the computational methods proposed in this work, may offer a useful approach in the modeling of retention of nucleoside and pterin compounds in HILIC.

Development of Gradient Retention Model in Ion Chromatography. Part II: Artificial Intelligence QSRR Approach

Methodology that integrates traditional QSRR modeling with transfer of information from isocratic to gradient environment was presented in the previous paper as an efficient new chromatographic approach. The previous research included application of conventional regression techniques that resulted in relatively high prediction errors. Also, it was shown that prediction error of the integrated model was mostly caused by prediction of the QSRR models. Therefore, artificial intelligence was applied in this work in order to improve the prediction ability of QSRR-based model: Artificial neural networks were selected for QSRR modeling, while genetic algorithm was used for the selection of optimal descriptors. Both artificial neural networks and genetic algorithm were optimized in order to build accurate and reliable QSRR models. Selection function, crossover function, and percentage of genes’ mutations were varied in case of genetic algorithm, while artificial neural networks were optimized by means of different network type, training algorithm and number of neurons in hidden layer. During retention modeling, basic QSRR models developed for specific eluent strength were upgraded to isocratic, and thereafter to gradient retention model. None of the three developed models showed systematic errors, and the obtained predictions (RMSEP 11.66, 10.67, and 7.10 %, respectively) indicated significant improvement from the results presented in previous paper.

Quantitative structure–retention relationships of ionic liquid cations in characterization of stationary phases for HPLC

In this report chemometric methodology was utilized in comprehensive column comparative characterization in order to define patterns of similarity and dissimilarity between stationary phases under investigation. A quantitative structure–retention relationship approach was adopted to obtain statistically significant molecular-descriptor-based models describing the retention mechanism of eight ionic liquids in two home-made and five commercially available chromatographic columns. Consequently their retention mechanism was also studied and discussed. Principal component analysis was adopted to reduce the dimensionality of experimental data in order to characterize different stationary phases based on the behavior of ionic liquids. QSRR models together with stationary phase characterization provided reliable information on the properties and characteristics of the studied columns.

2次元Quantitative Structure-Retention Relationshipモデルの構築と逆解析を利用した構造推定

2次元Quantitative Structure-Retention Relationshipモデルの構築と逆解析を利用した構造推定

Classification of LC columns based on the QSRR method and selectivity toward moclobemide and its metabolites

This paper focuses on a comparative study of the column classification system based on the quantitative structure-retention relationships (QSRR method) and column performance in real biomedical analysis. The assay was carried out for the LC separation of moclobemide and its metabolites in human plasma, using a set of 24 stationary phases. The QSRR models established for the studied stationary phases were compared with the column test performance results under two chemometric techniques – the principal component analysis (PCA) and the hierarchical clustering analysis (HCA). The study confirmed that the stationary phase classes found closely related by the QSRR approach yielded comparable separation for moclobemide and its metabolites. Therefore, the QSRR method could be considered supportive in the selection of a suitable column for the biomedical analysis offering the selection of similar or dissimilar columns with a relatively higher certainty.