| 1. |
- Brandmaier, Stefan, et al.
(författare)
-
PLS-Optimal: A stepwise D-Optimal design based on latent variables
- 2012
-
Ingår i: Journal of Chemical Information and Modeling. - : American Chemical Society (ACS). - 1549-9596 .- 1549-960X. ; 52:4, s. 975-983
-
Tidskriftsartikel (refereegranskat)abstract
- 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.
|
|
| 2. |
|
|
| 3. |
- Cassani, Stefano, et al.
(författare)
-
Evaluation of CADASTER QSAR Models for the Aquatic Toxicity of (Benzo)triazoles and Prioritisation by Consensus Prediction
- 2013
-
Ingår i: ATLA (Alternatives to Laboratory Animals). - : SAGE Publications. - 0261-1929. ; 41:1, s. 49-64
-
Tidskriftsartikel (refereegranskat)abstract
- QSAR regression models of the toxicity of triazoles and benzotriazoles ([B] TAZs) to an alga (Pseudokirchneriella subcapitata), Daphnia magna and a fish (Onchorhynchus mykiss), were developed by five partners in the FP7-EU Project, CADASTER. The models were developed by different methods - Ordinary Least Squares (OLS), Partial Least Squares (PLS), Bayesian regularised regression and Associative Neural Network (ASNN) - by using various molecular descriptors (DRAGON, PaDEL-Descriptor and QSPR-THESAURUS web). In addition, different procedures were used for variable selection, validation and applicability domain inspection. The predictions of the models developed, as well as those obtained in a consensus approach by averaging the data predicted from each model, were compared with the results of experimental tests that were performed by two CADASTER partners. The individual and consensus models were able to correctly predict the toxicity classes of the chemicals tested in the CADASTER project, confirming the utility of the QSAR approach. The models were also used for the prediction of aquatic toxicity of over 300 (B)TAZs, many of which are included in the REACH pre-registration list, and were without experimental data. This highlights the importance of QSAR models for the screening and prioritisation of untested chemicals, in order to reduce and focus experimental testing.
|
|
| 4. |
- Durjava, Mojca Kos, et al.
(författare)
-
Experimental Assessment of the Environmental Fate and Effects of Triazoles and Benzotriazole
- 2013
-
Ingår i: ATLA (Alternatives to Laboratory Animals). - : SAGE Publications. - 0261-1929. ; 41:1, s. 65-75
-
Tidskriftsartikel (refereegranskat)abstract
- The environmental fate and effects of triazoles and benzotriazoles are of concern within the context of chemical regulation. As part of an intelligent testing strategy, experimental tests were performed on endpoints that are relevant for risk assessment. The experimental tests included the assessment of eco-toxicity to an alga, a daphnid and zebrafish embryos, and the assessment of ready biodegradability. Triazole and benzotriazole compounds were selected for testing, based on existing toxicity data for vertebrate and invertebrate species, as well as on the principal component analysis of molecular descriptors aimed at selecting the minimum number of test compounds in order to maximise the chemical domain spanned for both compound classes. The experimental results show that variation in the toxicities of triazoles and benzotriazole across species was relatively minor; in general, the largest factor was approximately 20. The study conducted indicated that triazoles are not readily biodegradable.
|
|
| 5. |
- Golsteijn, Laura, et al.
(författare)
-
Assessing predictive uncertainty in comparative toxicity potentials of triazoles
- 2014
-
Ingår i: Environmental Toxicology and Chemistry. - : Wiley. - 0730-7268 .- 1552-8618. ; 33:2, s. 293-301
-
Tidskriftsartikel (refereegranskat)abstract
- Comparative toxicity potentials (CTPs) quantify the potential ecotoxicological impacts of chemicals per unit of emission. They are the product of a substance's environmental fate, exposure, and hazardous concentration. When empirical data are lacking, substance properties can be predicted. The goal of the present study was to assess the influence of predictive uncertainty in substance property predictions on the CTPs of triazoles. Physicochemical and toxic properties were predicted with quantitative structure-activity relationships (QSARs), and uncertainty in the predictions was quantified with use of the data underlying the QSARs. Degradation half-lives were based on a probability distribution representing experimental half-lives of triazoles. Uncertainty related to the species' sample size that was present in the prediction of the hazardous aquatic concentration was also included. All parameter uncertainties were treated as probability distributions, and propagated by Monte Carlo simulations. The 90% confidence interval of the CTPs typically spanned nearly 4 orders of magnitude. The CTP uncertainty was mainly determined by uncertainty in soil sorption and soil degradation rates, together with the small number of species sampled. In contrast, uncertainty in species-specific toxicity predictions contributed relatively little. The findings imply that the reliability of CTP predictions for the chemicals studied can be improved particularly by including experimental data for soil sorption and soil degradation, and by developing toxicity QSARs for more species. (c) 2013 SETAC
|
|
| 6. |
|
|
| 7. |
- Iqbal, Muhammad Sarfraz, et al.
(författare)
-
Treatment of Epistemic Uncertainty in Environmental Fate Models –Consequences on Chemical Safety Regulatory Strategies
- 2012
-
Konferensbidrag (övrigt vetenskapligt/konstnärligt)abstract
- The practical impact of treatment of epistemic uncertainty on decision making wasillustrated on two kinds of decisions from chemical regulation. First, regulatory strategies derivedfrom a simplified decision model based on toxicity and persistence showed that regulated level ofexposure is more conservative (safe) when uncertainty has been given a non-probabilistictreatment. Persistence and its uncertainty had been assessed by a Level II fugacity model forwhich input parameters had been quantified either by Bayesian probabilities, fuzzy numbers(non-probabilistic), or combinations of these (probability boxes). These findings are restricted tohow we let decision makers respond to uncertainty in model predictions by the chosen set ofdecision rules. Further, the use of either treatment depends on the quality and quantity ofbackground knowledge and the required level of detail on the assessment. In the absence ofexperimentally tested physicochemical endpoints, European chemical regulation REACH allowsthe use of non-testing strategies such as Quantitative Structure-Property Relationships (QSPR) topredict the required information. The second decision problem was to select which chemicalsubstances to prioritize for experimental testing in order to strengthen the background knowledgefor chemical regulation with respect to the uncertainty in QSPR predictions. We found that thevalue of reducing uncertainty, given by the expected gain in net benefit for society, was affectedby its treatment and there were no consistent order of testing of the three compounds. However,value of information is a Bayesian probabilistic approach that, unless developed further, loose itsinterpretability under other treatments of uncertainty. The framework of a predictive model, riskmodel, decision model and value of information analysis provides a computational template forfurther evaluation of the effect of treatment of uncertainty on decision making.
|
|
| 8. |
- Iqbal, Muhammad Sarfraz, et al.
(författare)
-
Understanding quantitative structure-property relationships uncertainty in environmental fate modeling
- 2013
-
Ingår i: Environmental Toxicology and Chemistry. - : Wiley. - 0730-7268 .- 1552-8618. ; 32:5, s. 1069-1076
-
Tidskriftsartikel (refereegranskat)abstract
- 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.
|
|
| 9. |
- Sahlin, Ullrika, et al.
(författare)
-
A benefit analysis of screening for invasive species - base-rate uncertainty and the value of information
- 2011
-
Ingår i: Methods in Ecology and Evolution. - 2041-210X. ; 2:5, s. 500-508
-
Tidskriftsartikel (refereegranskat)abstract
- 1.. Implementation of the full spectra of screening tools to prevent the introduction of invasive species results in a need to evaluate the cost-efficiency of gathering the information needed to screen for these species. 2. We show how the Bayesian value of information approach can be used to derive the benefit of a screening model based on species traits, which in combination with the base rate of invasiveness, i.e. the proportion of invasive species among those introduced and established, predicts species-specific invasiveness. 3. A pre-posterior Bayesian analysis demonstrated that the benefit of the screening model of invasiveness depends on both the accuracy in predictions and the uncertainty in the base rate of invasiveness. However, even though increasing model accuracy always generates higher model benefit, acknowledging or neglecting the uncertainty in the base rate of invasiveness does not. This means that uncertainty in the base rate is important to consider in the cost-benefit analysis of the screening model. 4. As an example, we derived the benefit of basing decisions on a screening model trained for a data set on species traits of invasive and non-invasive marine macroalgae introduced into Europe. The benefit ranged from 0.6% to 19% of the loss of introducing an invasive species, where the actual value can be estimated if we know the monetary values of impacts from introducing invasive and not introducing non-invasive species. 5. Cost-benefit analyses of screening models for invasive species is one means to reach efficient management of the risks of non-indigenous species. Value of information is a useful tool for benefit analysis of predictive models with respect to decision-making, which goes beyond the investigations of model accuracy. Here, we use value of information analysis to evaluate which sources of uncertainty that is most worth while to reduce and how to set the cost of gathering further species-specific information which will improve the accuracy of a screening.
|
|
| 10. |
- Sahlin, Ullrika, et al.
(författare)
-
A risk assessment perspective of current practice in characterizing uncertainties in QSAR regression predictions
- 2011
-
Ingår i: Molecular Informatics. - : Wiley. - 1868-1751 .- 1868-1743. ; 30:6-7, s. 551-564
-
Tidskriftsartikel (refereegranskat)abstract
- 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.
|
|
