HighlightsBlood DNA methylation profiles of 1156 individuals were assessed for age correlation.Stepwise regression identified 23 age-associated CpG sites in DNA from blood.A machine learning model based on 16 markers predicted age with a mean error of 3.8 years.The model predicted age successfully for twins and ‘diseased’ individuals.A new NGS-based method was combined with machine learning for age prediction.

Pharmaceuticals have been considered ‘contaminants of emerging concern’ for more than 20 years. In that time, many laboratory studies have sought to identify hazard and assess risk in the aquatic environment, whilst field studies have searched for targeted candidates and occurrence trends using advanced analytical techniques. However, a lack of a systematic approach to the detection and quantification of pharmaceuticals has provided a fragmented literature of serendipitous approaches. Evaluation of the extent of the risk for the plethora of human and veterinary pharmaceuticals available requires the reliable measurement of trace levels of contaminants across different environmental compartments (water, sediment, biota - of which biota has been largely neglected). The focus on pharmaceutical concentrations in surface waters and other exposure media have therefore limited both the characterisation of the exposome in aquatic wildlife and the understanding of cause and effect relationships. Here, we compile the current analytical approaches and available occurrence and accumulation data in biota to review the current state of research in the field. Our analysis provides evidence in support of the ‘Matthew Effect’ and raises critical questions about the use of targeted analyte lists for biomonitoring. We provide six recommendations to stimulate and improve future research avenues.

Abstract:The recent development of broad-scope high resolution mass spectrometry (HRMS) screening methods has resulted in a much improved capability for new compound identification in environmental samples. However, positive identifications at the ng/L concentration level rely on analytical reference standards for chromatographic retention time (t R ) and mass spectral comparisons. Chromatographic t R prediction can play a role in increasing confidence in suspect screening efforts for new compounds in the environment, especially when standards are not available, but reliable methods are lacking. The current work focuses on the development of artificial neural networks (ANNs) for t R prediction in gradient reversed-phase liquid chromatography and applied along with HRMS data to suspect screening of wastewater and environmental surface water samples. Based on a compound t R dataset of >500 compounds, an optimized 4-layer back-propagation multi-layer perceptron model enabled predictions for 85 % of all compounds to within 2 minutes of their measured t R for training (n=344) and verification (n=100) datasets. To evaluate the ANN ability for generalization to new data, the model was further tested using 100 randomly selected compounds and revealed 95 % prediction accuracy within the 2-minute elution interval. Given the increasing concern on the presence of drug metabolites and other transformation products (TPs) in the aquatic environment, the model was applied along with HRMS data for preliminary identification of pharmaceutically-related compounds in real samples. Examples of compounds where reference standards were subsequently acquired and later confirmed are also presented. To our knowledge, this work presents for the first time, the successful application of an accurate retention time predictor and HRMS data-mining using the largest number of compounds to preliminarily identify new or emerging contaminants in wastewater and surface waters.

The development, characterisation and application of a new analytical method for multi-residue PPCP determination in the freshwater amphipod, Gammarus pulex are presented. Analysis was performed using pulverised liquid extraction (PuLE), solid phase extraction (SPE) and liquid chromatography-tandem mass spectrometry (LC-MS/MS). Qualitative method performance offered excellent limits of detection at <20 ng g(-1) for 18 out of 29 compounds. For quantitative application, linearity and precision were considered acceptable for 10 compounds across the ng-μg g(-1) range (R2≥0.99; ≤20% relative standard deviation respectively). The method was applied to the analysis of G. pulex and river water sourced from six tributaries of the River Thames. Carbamazepine, diazepam, nimesulide, trimethoprim and warfarin were determined in G. pulex samples at low ng g(-1) (dry weight) concentrations across these sites. Temazepam and diclofenac were also detected, but were not quantifiable. Six pharmaceuticals were quantified in surface waters across the eight sites at concentrations ranging from 3 to 344 ng L(-1). The possibility for confirmatory detection and subsequent quantification of pharmaceutical residues in benthic organisms such as G. pulex will enable further understanding on the susceptibility and ecological effects of PPCPs in the aquatic environment.

The computational generation of gradient retention time data for retrospective detection of suspected sports doping species in postanalysis human urine sample data is presented herein. Retention data for a selection of 86 compounds included in the London 2012 Olympic and Paralympic Games drug testing schedule were used to train, verify, and test a range of computational models for this purpose. Spiked urine samples were analyzed using solid phase extraction followed by ultrahigh-pressure gradient liquid chromatography coupled to electrospray ionization high-resolution mass spectrometry. Most analyte retention times varied ≤0.2 min over the relatively short runtime of 10 min. Predicted retention times were within 0.5 min of experimental values for 12 out of 15 blind test compounds (largest error: 0.97 min). Minimizing the variance in predictive ability across replicate networks of identical architecture is presented for the first time along with a quantitative discussion of the contribution of each selected molecular descriptor toward the overall predicted value. The performance of neural computing predictions for isobaric compound retention time is also discussed. This work presents the application of neural networks to the prediction of gradient retention time in archived high-resolution urine analysis sample data for the first time in the field of anti-doping.

The modelling and prediction of reversed-phase chromatographic retention time (tR) under gradient elution conditions for 166 pharmaceuticals in wastewater extracts is presented using artificial neural networks for the first time. Radial basis function, multilayer perceptron and generalised regression neural networks were investigated and a comparison of their predictive ability for model solutions discussed. For real world application, the effect of matrix complexity on tR measurements is presented. Measured tR for some compounds in influent wastewater varied by >1min in comparison to tR in model solutions. Similarly, matrix impact on artificial neural network predictive ability was addressed towards developing a more robust approach for routine screening applications. Overall, the best neural network had a predictive accuracy of <1.3min at the 75th percentile of all measured tR data in wastewater samples (<10% of the total runtime). Coefficients of determination for 30 blind test compounds in wastewater matrices lay at or above R(2)=0.92. Finally, the model was evaluated for application to the semi-targeted identification of pharmaceutical residues during a weeklong wastewater sampling campaign. The model successfully identified native compounds at a rate of 83±4% and 73±5% in influent and effluent extracts, respectively. The use of an HRMS database and the optimised ANN model was also applied to shortlisting of 37 additional compounds in wastewater. Ultimately, this research will potentially enable faster identification of emerging contaminants in the environment through more efficient post-acquisition data mining.

This study considers whether the current standard toxicokinetic methods are an accurate and applicable assessment of xenobiotic exposure in an aquatic freshwater invertebrate. An in vivo exposure examined the uptake and elimination kinetics for eight pharmaceutical compounds in the amphipod crustacean, Gammarus pulex by measuring their concentrations in both biological material and in the exposure medium over a 96 h period. Selected pharmaceuticals included two anti-inflammatories (diclofenac and ibuprofen), two beta-blockers (propranolol and metoprolol), an anti-depressant (imipramine), an anti-histamine (ranitidine) and two beta-agonists (formoterol and terbutaline). Kinetic bioconcentration factors (BCFs) for the selected pharmaceuticals were derived from a first-order one-compartment model using either the simultaneous or sequential modelling methods. Using the simultaneous method for parameter estimation, BCF values ranged from 12 to 212. In contrast, the sequential method for parameter estimation resulted in bioconcentration factors ranging from 19 to 4533. Observed toxicokinetic plots showed statistically significant lack-of-fits and further interrogation of the models revealed a decreasing trend in the uptake rate constant over time for rantidine, diclofenac, imipramine, metoprolol, formoterol and terbutaline. Previous published toxicokinetic data for 14 organic micro-pollutants were also assessed and similar trends were identified to those observed in this study. The decreasing trend of the uptake rate constant over time highlights the need to interpret modelled data more comprehensively to ensure uncertainties associated with uptake and elimination parameters for determining bioconcentration factors are minimised.

The application of machine learning has recently gained interest from ecotoxicological fields for its ability to model and predict chemical and/or biological processes, such as the prediction of bioconcentration. However, comparison of different models and the prediction of bioconcentration in invertebrates has not been previously evaluated. A comparison of 24 linear and machine learning models is presented herein for the prediction of bioconcentration in fish and important factors that influenced accumulation identified. R and root mean square error (RMSE) for the test data (n = 110 cases) ranged from 0.23-0.73 and 0.34-1.20, respectively. Model performance was critically assessed with neural networks and tree-based learners showing the best performance. An optimised 4-layer multi-layer perceptron (14 descriptors) was selected for further testing. The model was applied for cross-species prediction of bioconcentration in a freshwater invertebrate, Gammarus pulex. The model for G. pulex showed good performance with R of 0.99 and 0.93 for the verification and test data, respectively. Important molecular descriptors determined to influence bioconcentration were molecular mass (MW), octanol-water distribution coefficient (logD), topological polar surface area (TPSA) and number of nitrogen atoms (nN) among others. Modelling of hazard criteria such as PBT, showed potential to replace the need for animal testing. However, the use of machine learning models in the regulatory context has been minimal to date and is critically discussed herein. The movement away from experimental estimations of accumulation to in silico modelling would enable rapid prioritisation of contaminants that may pose a risk to environmental health and the food chain.