Critical evaluation of a simple retention time predictor based on LogKow as a complementary tool in the identification of emerging contaminants in water

Abstract

Abstract:There has been great interest in environmental analytical chemistry in developing screening methods based on liquid chromatography-high resolution mass spectrometry (LC-HRMS) for emerging contaminants. Using HRMS, compound identification relies on the high mass resolving power and mass accuracy attainable by these analyzers.When dealing with wide-scope screening, retention time prediction can be a complementary tool for the identification of compounds, and can also reduce tedious data processing when several peaks appear in the extracted ion chromatograms. There are many in silico, Quantitative Structure-Retention Relationship methods available for the prediction of retention time for LC. However, most of these methods use commercial software to predict retention time based on various molecular descriptors. This paper explores the applicability and makes a critical discussion on a far simpler and cheaper approach to predict retention times by using LogKow. The predictor was based on a database of 595 compounds, their respective LogKow values and a chromatographic run time of 18 minutes. Approximately 95% of the compounds were found within 4.0 minutes of their actual retention times, and 70% within 2.0 minutes. A predictor based purely on pesticides was also made, enabling 80% of these compounds to be found within 2.0 minutes of their actual retention times. To demonstrate the utility of the predictors, they were successfully used as an additional tool in the identification of 30 commonly found emerging contaminants in water. Furthermore, a comparison was made by using different mass extraction windows to minimize the number of false positives obtained.

Suspect screening of large numbers of emerging contaminants in environmental waters using artificial neural networks for chromatographic retention time prediction and high resolution mass spectrometry data analysis

“…The coefficients of determination (R 2 ) were between 0.86 and 0.90 between the three sets, which was already a marked improvement on that obtained from our previous study at R 2 = 0.67. Furthermore, the root-mean-squared error (RMSE) of the blind set of compounds (1.03 min) is less than half that of our previous work (2.19 min) (Bade et al, 2015).…”

“…In a previous study, we developed a simple t R prediction model based on logK ow of nearly 600 compounds, predicted using freeware (Bade et al, 2015). This resulted in approximately 70 % of all compounds being predicted within 2 minutes of the measured t R , and 95 % within 4 minutes.…”

“…These covered a large range of molecular hydrophobicity (logK ow -3 to 9). Information relating to 595 standards was available (Bade et al, 2015), however after transforming the compounds using SMILES codes, some errors were observed, leading to incomplete data, and a further 42 were removed from the initial ANN method development (Section 3.1) to use in a subsequent blind test (Section 3.2) .Further details relating to these compounds can be found elsewhere (Bade et al, 2015;Hernández et al, 2015b).…”

“…It is frequently used in environmental studies for the description of the fate of various pollutants and as a simple t R predictor for TPs (Kern et al, 2009) and emerging contaminants (Bade et al, 2015;Nurmi et al, 2012). Alongside simple algorithms, other and more complex in silico approaches now exist which are based on quantitative structure-retention relationship (QSRR) modeling, including artificial neural networks, support vector machines and random forests (Giaginis and Tsantili-Kakoulidou, 2012;Héberger, 2007).…”

“…To the best of our knowledge, this method includes the largest range of physicochemically diverse compounds for this purpose (n=544 in total) and includes both neutral and charged compounds eluted under gradient reversed-phase LC conditions. Lastly, this work aimed to improve upon a recent logK ow -based t R prediction approach (Bade et al, 2015) using the ANNs as an alternative. This work, for the first time, presents the use of ANN for identification of additional suspect compounds (including metabolites and TPs) in wastewater and surface water samples both with and without reference standards.…”