Multi-way partial least squares modeling of water quality data

Chemometrics as a valuable tool for evaluating interactions between antiretroviral drugs and food

AIMS

Clinically significant interactions with food occur for more than half of antiretroviral drugs. Different physiochemical properties deriving from the chemical structures of antiretroviral drugs may contribute to the variable food effect. Chemometric methods allow analysing a large number of interrelated variables concomitantly and visualizing correlations between them. We used a chemometric approach to determine the types of correlations among different features of antiretroviral drugs and food that may influence interactions.

METHODS

Thirty-three antiretroviral drugs were analysed: ten nucleoside reverse transcriptase inhibitors, six non-nucleoside reverse transcriptase inhibitors, five integrase strand transfer inhibitors, ten protease inhibitors, one fusion inhibitor and one HIV maturation inhibitor. Input data for the analysis were collected from already published clinical studies, chemical records and calculations. We constructed a hierarchical partial least squares (PLS) model with three response parameters: postprandial change of time to reach maximum drug concentration (ΔTmax ), albumin binding (%) and logarithm of partition coefficient (logP). Predictor parameters were the first two principal components of principal component analysis (PCA) models for six groups of molecular descriptors.

RESULTS

PCA models explained 64.4% to 83.4% of the variance of the original parameters (average: 76.9%), whereas the PLS model had four significant components and explained 86.2% and 71.4% of the variance in the sets of predictor and response parameters, respectively. We observed 58 significant correlations between ΔTmax , albumin binding (%), logP and constitutional, topological, hydrogen bonding and charge-based molecular descriptors.

CONCLUSIONS

Chemometrics is a useful and valuable tool for analysing interactions between antiretroviral drugs and food.

Human Health Risk Assessment of Trace Elements in Tap Water and the Factors Influencing Its Value

(1) Background: The influence of tap water fittings construction and internal pipe-work on the release of heavy metals was investigated. (2) Methods: A statistical approach was applied for the examination of the chemistry of tap water in five different cities in southern Poland. In total, 500 samples were collected (from 100 to 101 samples in each city). The sampling protocol included information on the construction of the water supply network and the physicochemical parameters of measured tap water. (3) Results: The statistical analysis allowed to extract the crucial factors that affect the concentrations of trace elements in tap water. Age of connection, age of tap, age of pipe-work as well as material of connection, material of pipe-work and material of appliance reveal the most significant variability of concentrations observed for As, Al, Cd, Cu, Fe, Mn, Pb, and Zn. Calculated cancer risks (CRs) decrease with the following order of analysed elements Ni > Cd > Cr > As = Pb and can be associated with the factors that affect the appearance of such elements in tap water. The hazard index (HI) was evaluated as negligible in 59.1% of the sampling points and low in 40.1% for adults. For children, a high risk was observed in 0.2%, medium in 9.0%, negligible in 0.4%, and low for the rest of the analysed samples.

Effectiveness of groundwater heavy metal pollution indices studies by deep-learning

Globally, groundwater heavy metal (HM) pollution is a serious concern, threatening drinking water safety as well as human and animal health. Therefore, evaluation of groundwater HM pollution is essential to prevent accompanying hazardous ecological impacts. In this aspect, the effectiveness of various groundwater HM pollution evaluation approaches should be examined for their level of trustworthiness. In this study, 226 groundwater samples from Arang of Chhattisgarh state, India, were collected and analyzed. Measured concentration for various HMs were further used to calculate six groundwater pollution indices, such as the HM pollution index (HPI), HM evaluation index (HEI), contamination index (CI), entropy-weight based HM contamination index (EHCI), Heavy metal index (HMI), and principal component analysis-based metal index (PMI). Groundwater in the study area was mainly contaminated by elevated Cd, Fe, and Pb concentrations due to natural and anthropogenic pollution. Moreover, this study explored the performance of deep learning (DL)-based predictive models via comparative study. Two hidden layers with 26 and 19 neurons in the first and second hidden layers, respectively, were optimised along with rectified linear unit activation function. A mini-batch gradient descent was also applied to ensure smooth convergence of the training dataset into the model. Results demonstrated that the DL-PMI scored lowest errors, 0.022 for mean square error (MSE), 0.140 for mean absolute error (MAE), and 0.148 for root mean square error (RMSE), in the model validation than the other DL-based groundwater HM pollution model. Prediction performances of all pollution indices were also verified using artificial neural network (ANN)-based models, which also highlighted the lowest validation error for ANN-PMI (MSE = 3.93, MAE = 1.38, and RMSE = 1.98). Furthermore, the prediction accuracies of PMI using both ANN and DL models scored the highest R² value of 0.95 and 0.99, respectively. Therefore it is suggested that groundwater HM pollution using PMI as the best indexing approach in the present study area. Moreover, compared to benchmark, ANN, the DL performed better; hence, it could be concluded that the proposed DL model may be suitable approach in the field of computational chemistry by handling overfitting problems.

Chemometrics for environmental monitoring: a review

Environmental monitoring is necessary to ensure the overall health and conservation of an ecosystem. However, ecosystems (e.g. air, water, soil), are complex, involving numerous processes (both native and external), inputs, contaminants, and living organisms. As such, monitoring an environmental system is not a trivial task. The data obtained from natural systems is often multifaceted and convoluted, as a multitude of inputs can be intertwined within the matrix of the information obtained as part of a study. This means that trends and important results can be easily overlooked by conventional and single dimensional data analysis protocols. Recently, chemometric methods have emerged as a powerful method for maximizing the details contained within a chemical data set. Specifically, chemometrics refers to the use of mathematical and statistical analysis methods to evaluate chemical data, beyond univariant analysis. This type of analysis can provide a quantitative description of environmental measurements, while also having the capacity to reveal previously overlooked trends in data sets. Applying chemometrics to environmental data allows us to identify and describe the inter-relationship of environmental drivers, sources of contamination, and their potential impact upon the environment. This review aims to provide a detailed understanding of chemometric techniques, how they are currently used in environmental monitoring, and how these techniques can be used to improve current practices. An enhanced ability to monitor environmental conditions and to predict trends would be greatly beneficial to government and research agencies in their ability to develop environmental policies and analytical procedures.

Explaining and modeling the concentration and loading of Escherichia coli in a stream-A case study

Escherichia coli (E. coli) level in streams is a public health indicator. Therefore, being able to explain why E. coli levels are sometimes high and sometimes low is important. Using citizen science data from Fall Creek in central NY we found that complementarily using principal component analysis (PCA) and partial least squares (PLS) regression provided insights into the drivers of E. coli and a mechanism for predicting E. coli levels, respectively. We found that stormwater, temperature/season and shallow subsurface flow are the three dominant processes driving the fate and transport of E. coli. PLS regression modeling provided very good predictions under stormwater conditions (R² = 0.85 for log (E. coli concentration) and R² = 0.90 for log (E. coli loading)); predictions under baseflow conditions were less robust. But, in our case, both E. coli concentration and E. coli loading were significantly higher under stormwater condition, so it is probably more important to predict high-flow E. coli hazards than low-flow conditions. Besides previously reported good indicators of in-stream E. coli level, nitrate-/nitrite-nitrogen and soluble reactive phosphorus were also found to be good indicators of in-stream E. coli levels. These findings suggest management practices to reduce E. coli concentrations and loads in-streams and, eventually, reduce the risk of waterborne disease outbreak.

Prediction of biochemical oxygen demand at the upstream catchment of a reservoir using adaptive neuro fuzzy inference system

The aim of this study is to examine the potential of adaptive neuro fuzzy inference system (ANFIS) to estimate biochemical oxygen demand (BOD). To illustrate the applicability of ANFIS method, the upstream catchment of Feitsui Reservoir in Taiwan is chosen as the case study area. The appropriate input variables used to develop the ANFIS models are determined based on the t-test. The results obtained by ANFIS are compared with those by multiple linear regression (MLR) and artificial neural networks (ANNs). Simulated results show that the identified ANFIS model is superior to the traditional MLR and nonlinear ANNs models in terms of the performance evaluated by the Pearson coefficient of correlation, the root mean square error, the mean absolute percentage, and the mean absolute error. These results indicate that ANFIS models are more suitable than ANNs or MLR models to predict the nonlinear relationship within the variables caused by the complexity of aquatic systems and to produce the best fit of the measured BOD concentrations. ANFIS can be seen as a powerful predictive alternative to traditional water quality modeling techniques and extended to other areas to improve the understanding of river pollution trends.

Chemometric analysis of the interactions among different parameters describing health conditions, breast cancer risk and fatty acids profile in serum of rats supplemented with conjugated linoleic acids

We investigated how different doses of conjugated linoleic acids applied for various periods of time influence breast cancer risk and fatty acids profile in serum of rats treated or not with 7,12-dimethylbenz[a]anthracene (DMBA). We also search for interactions among parameters describing health conditions and cancer risk. Animals were divided into 18 groups with different diet modifications (vegetable oil, 1.0%, 2.0% additions of CLA) and different periods of supplementation. In groups treated with DMBA mammary adenocarcinomas appeared. Due to the complexity of experiment apart from statistical analysis a chemometric tool-Partial Least Square method was applied. Analysis of pairs of correlated parameters allowed to identify some regularities concerning the relationships between fatty acid profiles and clinical features of animals. Fatty acids profile was the result of prolonged exposure to high dose of CLA and DMBA administration. These two factors underlined the differences in fatty acids profiles among clusters of animals.

Distribution of uranium, thorium and some stable trace and toxic elements in human hair and nails in Niška Banja Town, a high natural background radiation area of Serbia (Balkan Region, South-East Europe)

Human hair and nails can be considered as bio-indicators of the public exposure to certain natural radionuclides and other toxic metals over a long period of months or even years. The level of elements in hair and nails usually reflect their levels in other tissues of body. Niška Banja, a spa town located in southern Serbia, with locally high natural background radiation was selected for the study. To assess public exposure to the trace elements, hair and nail samples were collected and analyzed. The concentrations of uranium, thorium and some trace and toxic elements (Mn, Ni, Cu, Sr, Cd, and Cs) were determined using inductively coupled plasma mass spectrometry (ICP-MS). U and Th concentrations in hair varied from 0.0002 to 0.0771 μg/g and from 0.0002 to 0.0276 μg/g, respectively. The concentrations in nails varied from 0.0025 to 0.0447 μg/g and from 0.0023 to 0.0564 μg/g for U and Th, respectively. We found significant correlations between some elements in hair and nails. Also indications of spatial clustering of high values could be found. However, this phenomenon as well as the large variations in concentrations of heavy metals in hair and nail could not be explained. As hypotheses, we propose possible exposure pathways which may explain the findings, but the current data does not allow testing them.

Establishment of a structural equation model for ground-level ozone: a case study at an urban roadside site

This study established a cause-effect relationship between ground-level ozone and latent variables employing partial least-squares analysis at an urban roadside site in four distinct seasons. Two multivariate analytic methods, factor analysis, and cluster analysis were adopted to cite and identify suitable latent variables from 14 observed variables (i.e., meteorological factors, wind and primary air pollutants) in 2008-2010. Analytical results showed that the first six components explained 80.3 % of the variance, and eigenvalues of the first four components were greater than 1. The effectiveness of this model was empirically confirmed with three indicators. Except for surface pressure, factor loadings of observed variables were 0.303-0.910 and reached statistical significance at the 5 % level. Composite reliabilities for latent variables were 0.672-0.812 and average variances were 0.404-0.547, except for latent variable "primary" in spring; thus, discriminant validity and convergent validity were marginally accepted. The developed model is suitable for the assessment of urban roadside surface ozone, considering interactions among meteorological factors, wind factors, and primary air pollutants in each season.

Linear and nonlinear modeling approaches for urban air quality prediction

In this study, linear and nonlinear modeling was performed to predict the urban air quality of the Lucknow city (India). Partial least squares regression (PLSR), multivariate polynomial regression (MPR), and artificial neural network (ANN) approach-based models were constructed to predict the respirable suspended particulate matter (RSPM), SO(2), and NO(2) in the air using the meteorological (air temperature, relative humidity, wind speed) and air quality monitoring data (SPM, NO(2), SO(2)) of five years (2005-2009). Three different ANN models, viz. multilayer perceptron network (MLPN), radial-basis function network (RBFN), and generalized regression neural network (GRNN) were developed. All the five different models were compared for their generalization and prediction abilities using statistical criteria parameters, viz. correlation coefficient (R), standard error of prediction (SEP), mean absolute error (MAE), root mean squared error (RMSE), bias, accuracy factor (A(f)), and Nash-Sutcliffe coefficient of efficiency (E(f)). Nonlinear models (MPR, ANNs) performed relatively better than the linear PLSR models, whereas, performance of the ANN models was better than the low-order nonlinear MPR models. Although, performance of all the three ANN models were comparable, the GRNN over performed the other two variants. The optimal GRNN models for RSPM, NO(2), and SO(2) yielded high correlation (between measured and model predicted values) of 0.933, 0.893, and 0.885; 0.833, 0.602, and 0.596; and 0.932, 0.768 and 0.729, respectively for the training, validation and test sets. The sensitivity analysis performed to evaluate the importance of the input variables in optimal GRNN revealed that SO(2) was the most influencing parameter in RSPM model, whereas, SPM was the most important input variable in other two models. The ANN models may be useful tools in the air quality predictions.

Iodine, selenium, and other trace elements in urine of pregnant women

The purpose of this work was to determine trace element levels in urine and evaluate possible associations between urinary iodine concentration (UIC), other trace elements (Cr, Cu, Fe, Mn, Na, Se, Zn), toxic elements (Cd, Pb), anthropometrical measures (body weight and height), glycemic indices (serum insulin and glucose), and several parameters related to thyroid function (thyroid stimulating hormone, free thyroxine, antithyroid peroxidase antibodies, thyroid volume, and thyroid echogenicity) in pregnant women. One hundred sixty-nine participants were recruited. The whole study group, originating from Krakow region, comprised three subgroups belonging to three trimesters: I trimester (n = 28), II trimester (n = 83), and III trimester (n = 58). Trace elements were determined using inductively coupled plasma mass/(atomic emission) spectrometry. Partial least square model was used to reveal correlation structure between parameters investigated, as well as a possible causal relationship between dependent parameters and potentially explanatory parameters. Results obtained for trace and toxic elements in urine were comparable with results of other authors, although the study group was not homogenous. We confirmed (1) low iodine excretion in pregnant women, (2) the existence of statistically significant correlation between UIC and urinary selenium, and (3) lack of correlation between latter parameter and typical indices of thyroid function. Urinary selenium correlated with other urinary trace elements, but physiological significance of this finding remains uncertain. The fact that a large number of pregnant women fail to meet dietary recommendations for iodine is the major reason for concern.

Chemometrics assisted spectrophotometric determination of pyridine in water and wastewater

The paper reports a direct method for the determination of pyridine in water and wastewater samples based on ultraviolet spectrophotometric measurements using multi-way modeling techniques. Parallel factor analysis (PARAFAC) and multi-way partial least squares (N-PLS) regression methods were employed for the decomposition of spectra and quantification of pyridine. The study was carried out in the pH range of 1.0-12.0 and concentration range of 0.67-51.7 microgmL(-1) of pyridine. Both the three-way PARAFAC and tri-PLS1 models successfully predicted the concentration of pyridine in synthetic (spiked) river water and field wastewater samples. The mean recovery obtained from PARAFAC regression model were 97.39% for the spiked and 99.84% for the field wastewater samples, respectively. The sensitivity and precision of the method for pyridine determination were 0.58% and 5.95%, respectively. The N-PLS regression model yielded mean recoveries of 99.29% and 100.18% for the spiked and field wastewater samples, respectively. The prediction accuracy of the methods was evaluated through the root mean square error of prediction (RMSEP). For PARAFAC, it was 0.65 and 0.82 microgmL(-1) for spiked river water and field wastewater samples, respectively, while for N-PLS, it was 0.25 and 0.37 microgmL(-1), respectively. Both the PARAFAC and N-PLS methods, thus, yielded satisfactory results for the prediction of pyridine concentration in water and wastewater samples.