The QSPR-THESAURUS: the online platform of the CADASTER project.

The aim of the CADASTER project (CAse Studies on the Development and Application of in Silico Techniques for Environmental Hazard and Risk Assessment) was to exemplify REACH-related hazard assessments for four classes of chemical compound, namely, polybrominated diphenylethers, per and polyfluorinated compounds, (benzo)triazoles, and musks and fragrances. The QSPR-THESAURUS website (http: / /qspr-thesaurus.eu) was established as the project's online platform to upload, store, apply, and also create, models within the project. We overview the main features of the website, such as model upload, experimental design and hazard assessment to support risk assessment, and integration with other web tools, all of which are essential parts of the QSPR-THESAURUS.

Understanding quantitative structure-property relationships uncertainty in environmental fate modeling.

In cases in which experimental data on chemical-specific input parameters are lacking, chemical regulations allow the use of alternatives to testing, such as in silico predictions based on quantitative structure-property relationships (QSPRs). Such predictions are often given as point estimates; however, little is known about the extent to which uncertainties associated with QSPR predictions contribute to uncertainty in fate assessments. In the present study, QSPR-induced uncertainty in overall persistence (POV ) and long-range transport potential (LRTP) was studied by integrating QSPRs into probabilistic assessments of five polybrominated diphenyl ethers (PBDEs), using the multimedia fate model Simplebox. The uncertainty analysis considered QSPR predictions of the fate input parameters' melting point, water solubility, vapor pressure, organic carbon-water partition coefficient, hydroxyl radical degradation, biodegradation, and photolytic degradation. Uncertainty in POV and LRTP was dominated by the uncertainty in direct photolysis and the biodegradation half-life in water. However, the QSPRs developed specifically for PBDEs had a relatively low contribution to uncertainty. These findings suggest that the reliability of the ranking of PBDEs on the basis of POV and LRTP can be substantially improved by developing better QSPRs to estimate degradation properties. The present study demonstrates the use of uncertainty and sensitivity analyses in nontesting strategies and highlights the need for guidance when compounds fall outside the applicability domain of a QSPR.

Description and propagation of uncertainty in input parameters in environmental fate models.

Today, chemical risk and safety assessments rely heavily on the estimation of environmental fate by models. The key compound-related properties in such models describe partitioning and reactivity. Uncertainty in determining these properties can be separated into random and systematic (incompleteness) components, requiring different types of representation. Here, we evaluate two approaches that are suitable to treat also systematic errors, fuzzy arithmetic, and probability bounds analysis. When a best estimate (mode) and a range can be computed for an input parameter, then it is possible to characterize the uncertainty with a triangular fuzzy number (possibility distribution) or a corresponding probability box bound by two uniform distributions. We use a five-compartment Level I fugacity model and reported empirical data from the literature for three well-known environmental pollutants (benzene, pyrene, and DDT) as illustrative cases for this evaluation. Propagation of uncertainty by discrete probability calculus or interval arithmetic can be done at a low computational cost and gives maximum flexibility in applying different approaches. Our evaluation suggests that the difference between fuzzy arithmetic and probability bounds analysis is small, at least for this specific case. The fuzzy arithmetic approach can, however, be regarded as less conservative than probability bounds analysis if the assumption of independence is removed. Both approaches are sensitive to repeated parameters that may inflate the uncertainty estimate. Uncertainty described by probability boxes was therefore also propagated through the model by Monte Carlo simulation to show how this problem can be avoided.

Assessing bioaccumulation of polybrominated diphenyl ethers for aquatic species by QSAR modeling.

Polybrominated diphenyl ethers (PBDEs) are used as flame retardants in textiles, foams and plastics. Highly bioaccumulative with toxic effects including developmental neurotoxicity estrogen and thyroid hormones disruption, they are considered as persistent organic pollutants (POPs) and have been found in human tissues, wildlife and biota worldwide. But only some of them are banned from EU market. For the environmental fate studies of these compounds the bioconcentration factor (BCF) is one of the most important endpoints to start with. We applied quantitative structure-activity relationships techniques to overcome the limited experimental data and avoid more animal testing. The aim of this work was to assess the bioaccumulation of PBDEs by means of QSAR. First, a BCF dataset of specifically conducted experiments was modeled. Then the study was extended by predicting the bioaccumulation and biomagnification factors using some experimental values from the literature. Molecular descriptors were calculated using DRAGON 6. The most relevant ones were selected and resulting models were compared paying attention to the applicability domain.

PLS-optimal: a stepwise D-optimal design based on latent variables.

Several applications, such as risk assessment within REACH or drug discovery, require reliable methods for the design of experiments and efficient testing strategies. Keeping the number of experiments as low as possible is important from both a financial and an ethical point of view, as exhaustive testing of compounds requires significant financial resources and animal lives. With a large initial set of compounds, experimental design techniques can be used to select a representative subset for testing. Once measured, these compounds can be used to develop quantitative structure-activity relationship models to predict properties of the remaining compounds. This reduces the required resources and time. D-Optimal design is frequently used to select an optimal set of compounds by analyzing data variance. We developed a new sequential approach to apply a D-Optimal design to latent variables derived from a partial least squares (PLS) model instead of principal components. The stepwise procedure selects a new set of molecules to be measured after each previous measurement cycle. We show that application of the D-Optimal selection generates models with a significantly improved performance on four different data sets with end points relevant for REACH. Compared to those derived from principal components, PLS models derived from the selection on latent variables had a lower root-mean-square error and a higher Q2 and R2. This improvement is statistically significant, especially for the small number of compounds selected.

The chemical and environmental property space of REACH chemicals.

The European regulation on chemicals, REACH (Registration, Evaluation, Authorisation and Restriction of Chemicals), came into force on 1 June 2007. With pre-registration complete in 2008, data for these substances may provide an overview of the expected chemical space and its characteristics. In this paper, using various in silico computation tools, we evaluate 48782 neutral organic compounds from the list to identify hazardous and safe compounds. Two different classification schemes (modified Verhaar and ECOSAR) identified between 17% and 25% of the compounds as expressing only baseline toxicity (narcosis). A smaller portion could be identified as reactive (19%) or specifically acting (2.7%), while the majority were non-assigned (61%). Overall environmental persistence, bioaccumulation and long-range transport potential were evaluated using structure-activity relationships and a multimedia fugacity-based model. A surprisingly high proportion of compounds (20%), mainly aromatic and halogenated, had a very high estimated persistence (>195 d). The proportion of compounds with a very high estimated bioconcentration or bioaccumulation factor (>5000) was substantially less (6.9%). Finally, a list was compiled of those compounds within the applicability domain of the models used, meeting both persistence and bioaccumulation criteria, and with a long-range transport potential comparable to PCB. This list of 68 potential persistent organic pollutants contained many well-known compounds (all halogenated), but notably also five fluorinated compounds that were not included in the EINECS inventory. This study demonstrates the usability of in silico tools for identification of potentially environmentally hazardous chemicals.

Climate change - An uncertainty factor in risk analysis of contaminated land.

Metals frequently occur at contaminated sites, where their potential toxicity and persistence require risk assessments that consider possible long-term changes. Changes in climate are likely to affect the speciation, mobility, and risks associated with metals. This paper provides an example of how the climate effect can be inserted in a commonly used exposure model, and how the exposure then changes compared to present conditions. The comparison was made for cadmium (Cd) exposure to 4-year-old children at a highly contaminated iron and steel works site in southeastern Sweden. Both deterministic and probabilistic approaches (through probability bounds analysis, PBA) were used in the exposure assessment. Potential climate-sensitive variables were determined by a literature review. Although only six of the total 39 model variables were assumed to be sensitive to a change in climate (groundwater infiltration, hydraulic conductivity, soil moisture, soil:water distribution, and two bioconcentration factors), the total exposure was clearly affected. For example, by altering the climate-sensitive variables in the order of 15% to 20%, the deterministic estimate of exposure increased by 27%. Similarly, the PBA estimate of the reasonable maximum exposure (RME, defined as the upper bound of the 95th percentile) increased by almost 20%. This means that sites where the exposure in present conditions is determined to be slightly below guideline values may in the future exceed these guidelines, and risk management decisions could thus be affected. The PBA, however, showed that there is also a possibility of lower exposure levels, which means that the changes assumed for the climate-sensitive variables increase the total uncertainty in the probabilistic calculations. This highlights the importance of considering climate as a factor in the characterization of input data to exposure assessments at contaminated sites. The variable with the strongest influence on the result was the soil:water distribution coefficient (Kd).

Comparison of theoretical and experimental models for characterizing solvent properties using reversed phase liquid chromatography.

Theoretical and experimental quantitative structure-retention relationships (QSRR) models are useful for characterizing solvent properties and column selectivity in reversed phase liquid chromatography (RPLC). The chromatographic behavior of a model analyte, the herbicide atrazine, in a system derived from nine organic solvents and three chromatographic columns was used for developing QSRR models. Multiple linear regression (MLR) and partial least squares regression (PLSR) were used as statistical approaches. The similarities and differences between linear solvation energy relationships (LSER), and semi-empirical and theoretical molecular models were demonstrated. QSRR models show high predictive power, and can successfully predict retention factor (log k) for new solvents. The models are useful for solvent optimization and reducing time for method development in RPLC. The herbicide atrazine can be readily analyzed at a low level, and all three columns provided good resolution, high-performance and symmetrical peaks. The method is suitable for analysis of atrazine in water samples.

CADASTER QSPR Models for Predictions of Melting and Boiling Points of Perfluorinated Chemicals.

Quantitative structure property relationship (QSPR) studies on per- and polyfluorinated chemicals (PFCs) on melting point (MP) and boiling point (BP) are presented. The training and prediction chemicals used for developing and validating the models were selected from Syracuse PhysProp database and literatures. The available experimental data sets were split in two different ways: a) random selection on response value, and b) structural similarity verified by self-organizing-map (SOM), in order to propose reliable predictive models, developed only on the training sets and externally verified on the prediction sets. Individual linear and non-linear approaches based models developed by different CADASTER partners on 0D-2D Dragon descriptors, E-state descriptors and fragment based descriptors as well as consensus model and their predictions are presented. In addition, the predictive performance of the developed models was verified on a blind external validation set (EV-set) prepared using PERFORCE database on 15 MP and 25 BP data respectively. This database contains only long chain perfluoro-alkylated chemicals, particularly monitored by regulatory agencies like US-EPA and EU-REACH. QSPR models with internal and external validation on two different external prediction/validation sets and study of applicability-domain highlighting the robustness and high accuracy of the models are discussed. Finally, MPs for additional 303 PFCs and BPs for 271 PFCs were predicted for which experimental measurements are unknown.

A Risk Assessment Perspective of Current Practice in Characterizing Uncertainties in QSAR Regression Predictions.

The European REACH legislation accepts the use of non-testing methods, such as QSARs, to inform chemical risk assessment. In this paper, we aim to initiate a discussion on the characterization of predictive uncertainty from QSAR regressions. For the purpose of decision making, we discuss applications from the perspective of applying QSARs to support probabilistic risk assessment. Predictive uncertainty is characterized by a wide variety of methods, ranging from pure expert judgement based on variability in experimental data, through data-driven statistical inference, to the use of probabilistic QSAR models. Model uncertainty is dealt with by assessing confidence in predictions and by building consensus models. The characterization of predictive uncertainty would benefit from a probabilistic formulation of QSAR models (e.g. generalized linear models, conditional density estimators or Bayesian models). This would allow predictive uncertainty to be quantified as probability distributions, such as Bayesian predictive posteriors, and likelihood-based methods to address model uncertainty. QSAR regression models with point estimates as output may be turned into a probabilistic framework without any loss of validity from a chemical point of view. A QSAR model for use in probabilistic risk assessment needs to be validated for its ability to make reliable predictions and to quantify associated uncertainty.

Variability and uncertainty in Swedish exposure factors for use in quantitative exposure assessments.

Information of exposure factors used in quantitative risk assessments has previously been compiled and reported for U.S. and European populations. However, due to the advancement of science and knowledge, these reports are in continuous need of updating with new data. Equally important is the change over time of many exposure factors related to both physiological characteristics and human behavior. Body weight, skin surface, time use, and dietary habits are some of the most obvious examples covered here. A wealth of data is available from literature not primarily gathered for the purpose of risk assessment. Here we review a number of key exposure factors and compare these factors between northern Europe--here represented by Sweden--and the United States. Many previous compilations of exposure factor data focus on interindividual variability and variability between sexes and age groups, while uncertainty is mainly dealt with in a qualitative way. In this article variability is assessed along with uncertainty. As estimates of central tendency and interindividual variability, mean, standard deviation, skewness, kurtosis, and multiple percentiles were calculated, while uncertainty was characterized using 95% confidence intervals for these parameters. The presented statistics are appropriate for use in deterministic analyses using point estimates for each input parameter as well as in probabilistic assessments.

Applicability domains for classification problems: Benchmarking of distance to models for Ames mutagenicity set.

The estimation of accuracy and applicability of QSAR and QSPR models for biological and physicochemical properties represents a critical problem. The developed parameter of "distance to model" (DM) is defined as a metric of similarity between the training and test set compounds that have been subjected to QSAR/QSPR modeling. In our previous work, we demonstrated the utility and optimal performance of DM metrics that have been based on the standard deviation within an ensemble of QSAR models. The current study applies such analysis to 30 QSAR models for the Ames mutagenicity data set that were previously reported within the 2009 QSAR challenge. We demonstrate that the DMs based on an ensemble (consensus) model provide systematically better performance than other DMs. The presented approach identifies 30-60% of compounds having an accuracy of prediction similar to the interlaboratory accuracy of the Ames test, which is estimated to be 90%. Thus, the in silico predictions can be used to halve the cost of experimental measurements by providing a similar prediction accuracy. The developed model has been made publicly available at http://ochem.eu/models/1 .

Land-use versus natural controls on soil fertility in the Subandean Amazon, Peru.

Deforestation to amplify the agricultural frontier is a serious threat to the Amazon forest. Strategies to attain and maintain satisfactory soil fertility, which requires knowledge of spatial and temporal changes caused by land-use, are important for reaching sustainable development. This study highlights these issues by evaluating the relative effects of agricultural land-use and natural factors on chemical fertility of Inceptisols on redbed lithologies in the Subandean Amazon. Macro and micronutrients were determined in topsoil and subsoil in the vicinity of two villages at a total of 80 sites including pastures, coffee plantations, swidden fields, secondary forest and, as a reference, adjacent primary forest. Differences in soil fertility between the land cover classes were investigated by principal component analysis (PCA) and partial least squares regression (PLSR). Primary forest soil was found to be chemically similar to that of coffee plantations, pastures and secondary forests. There were no significant differences between soils of these land cover types in terms of plant nutrients (e.g. N, P, K, Ca, Mg, Mo, Mn, Zn, Cu and Co) or other fertility indicators (OM, pH, BS, EC, CECe and exchangeable acidity). The parent material (as indicated by texture and sample geographical origin) and the slope of the sampled sites were stronger controls on soil fertility than land cover type. Elevated concentrations of a few nutrients (NO(3) and K) were, however detected in soils of swidden fields. Despite being fertile (higher CECe, Ca and P) compared to Oxisols and Ultisols in the Amazon lowland, the Subandean soils frequently showed deficiencies in several nutrients (e.g. P, K, NO(3), Cu and Zn), and high levels of free Al at acidic sites. This paper concludes that deforestation and agricultural land-use has not introduced lasting chemical changes in the studied Subandean soils that are significant in comparison to the natural variability.

Exposure to contaminated sediments during recreational activities at a public bathing place.

More and more time is spent on recreational activities, but few risk assessments focus specifically on these situations and exposure factor data are often scarce. To assess exposure to contaminants at a public bathing place in an urban environment, we have compiled literature data, conducted observation studies, and analyzed water and sediment samples. The levels of anthropogenic contaminants are high in urban environments and traffic frequently plays an important role. In this study, to characterize variability and uncertainty, the deterministic exposure calculations for metal pollutants were supplemented by a probability bounds analysis for the polycyclic aromatic hydrocarbons (PAH). The results from these calculations show that oral intake is the major exposure route for metals, while skin absorption, with present assumptions, is more important for the PAH. The presently measured levels of contaminants, at this public bathing place, cannot be anticipated to cause any significant adverse influence on public health. This assessment methodology is easy to adapt and can be used routinely in other situations with more heavily contaminated surface sediments and lake water.

Identity and dynamics of putative N2 -fixing picoplankton in the Baltic Sea proper suggest complex patterns of regulation.

Heterocystous filamentous cyanobacteria are regarded as the main N2 -fixing organisms (diazotrophs) in the Baltic Sea. However, some studies indicate that picoplankton may also be important. The aim of this study was to examine the composition of putative diazotrophs in the picoplankton (< 3 µm) and to identify links to environmental factors. Nitrogenase (nifH) genes were amplified from community DNA by nested PCR, followed by cloning and sequencing. Clone libraries from nine environmental samples collected from the central Baltic Sea (April-October 2003, 3 m depth) and a negative control yielded a total of 433 sequences with an average clone library coverage of 92%. The sequences fell within nifH Clusters I, II and III and formed 15 distinct groups (> 96% amino acid similarity). Most of the sequences (77%) fell into nifH Cluster I (cyanobacteria and α-, ß- and γ-Proteobacteria). However, only 26 sequences were related to cyanobacteria (e.g. Pseudanabaena) and among these no unicellular phylotypes were found. Sequences clustering with alternative nitrogenases (anfH) and Archaea were found in one sample while sequences related to anaerobic phylotypes were found in six samples distributed throughout the season. The identified phylogenetic groups showed covariance with several environmental factors but no strong links could be established. This suggests a variable and complex regulation of diazotrophic groups within Baltic Sea picoplankton.

Critical assessment of QSAR models of environmental toxicity against Tetrahymena pyriformis: focusing on applicability domain and overfitting by variable selection.

The estimation of the accuracy of predictions is a critical problem in QSAR modeling. The "distance to model" can be defined as a metric that defines the similarity between the training set molecules and the test set compound for the given property in the context of a specific model. It could be expressed in many different ways, e.g., using Tanimoto coefficient, leverage, correlation in space of models, etc. In this paper we have used mixtures of Gaussian distributions as well as statistical tests to evaluate six types of distances to models with respect to their ability to discriminate compounds with small and large prediction errors. The analysis was performed for twelve QSAR models of aqueous toxicity against T. pyriformis obtained with different machine-learning methods and various types of descriptors. The distances to model based on standard deviation of predicted toxicity calculated from the ensemble of models afforded the best results. This distance also successfully discriminated molecules with low and large prediction errors for a mechanism-based model developed using log P and the Maximum Acceptor Superdelocalizability descriptors. Thus, the distance to model metric could also be used to augment mechanistic QSAR models by estimating their prediction errors. Moreover, the accuracy of prediction is mainly determined by the training set data distribution in the chemistry and activity spaces but not by QSAR approaches used to develop the models. We have shown that incorrect validation of a model may result in the wrong estimation of its performance and suggested how this problem could be circumvented. The toxicity of 3182 and 48774 molecules from the EPA High Production Volume (HPV) Challenge Program and EINECS (European chemical Substances Information System), respectively, was predicted, and the accuracy of prediction was estimated. The developed models are available online at http://www.qspr.org site.

Conjoint analysis for environmental evaluation--a review of methods and applications.

BACKGROUND, AIM AND SCOPE: Conjoint analysis and the related choice-modelling methods have been used for many years in marketing research to evaluate consumer behaviour and preferences for different kinds of product attributes. Recently, the number of applications in environmental science and management has started to grow. Conjoint analysis is found in many different forms, and the environmental studies evaluated in this review display the same range of methods as in other fields. The key characteristic of all these methods is that trade-offs are evaluated by jointly considering a number of important attributes. MAIN FEATURES: This paper is a review of the literature on environmental applications of conjoint analysis and assesses in which environmental area conjoint analysis has been most successful. The method and the design of the studies are reviewed as well. RESULTS: A total of 84 studies were found, dealing with environmental issues that were evaluated by conjoint analysis. The studies concern agriculture, ecosystem management, energy, environmental evaluation, forestry, land management, pollution, products, recreation, environmental risk analysis and waste management. DISCUSSION: Choice experiments seem to have a comparatively stronger position in environmental studies than elsewhere. Most of the environmental applications are related to natural resource management. This is somewhat surprising, but a number of reports have appeared also on product evaluation, which could be a key application area in the future. CONCLUSIONS: Compared to marketing and transportation, the number of environmental conjoint studies is rather small but increasing, and the method has proven to work effectively in eliciting preferences on environmental issues. In environmental issues, experimenters often use choice experiments, especially concerning ecosystem management and environmental evaluations. When it comes to evaluating preferences concerning agriculture, forestry, energy and products, a more traditional approach of conjoint analysis is favoured. RECOMMENDATIONS AND PERSPECTIVES: Two new areas of application are identified in this review--environmental communication and expert elicitation. Conjoint analysis can thus be developed into a useful instrument for environmental risk analysis and communication, both of which are necessary for an efficient approach to risk governance.

Combinatorial QSAR modeling of chemical toxicants tested against Tetrahymena pyriformis.

Selecting most rigorous quantitative structure-activity relationship (QSAR) approaches is of great importance in the development of robust and predictive models of chemical toxicity. To address this issue in a systematic way, we have formed an international virtual collaboratory consisting of six independent groups with shared interests in computational chemical toxicology. We have compiled an aqueous toxicity data set containing 983 unique compounds tested in the same laboratory over a decade against Tetrahymena pyriformis. A modeling set including 644 compounds was selected randomly from the original set and distributed to all groups that used their own QSAR tools for model development. The remaining 339 compounds in the original set (external set I) as well as 110 additional compounds (external set II) published recently by the same laboratory (after this computational study was already in progress) were used as two independent validation sets to assess the external predictive power of individual models. In total, our virtual collaboratory has developed 15 different types of QSAR models of aquatic toxicity for the training set. The internal prediction accuracy for the modeling set ranged from 0.76 to 0.93 as measured by the leave-one-out cross-validation correlation coefficient ( Q abs2). The prediction accuracy for the external validation sets I and II ranged from 0.71 to 0.85 (linear regression coefficient R absI2) and from 0.38 to 0.83 (linear regression coefficient R absII2), respectively. The use of an applicability domain threshold implemented in most models generally improved the external prediction accuracy but at the same time led to a decrease in chemical space coverage. Finally, several consensus models were developed by averaging the predicted aquatic toxicity for every compound using all 15 models, with or without taking into account their respective applicability domains. We find that consensus models afford higher prediction accuracy for the external validation data sets with the highest space coverage as compared to individual constituent models. Our studies prove the power of a collaborative and consensual approach to QSAR model development. The best validated models of aquatic toxicity developed by our collaboratory (both individual and consensus) can be used as reliable computational predictors of aquatic toxicity and are available from any of the participating laboratories.