Application of Bayesian networks for hazard ranking of nanomaterials to support human health risk assessment

Marvin, H.J.P.; Bouzembrak, Yamine; Janssen, Esmée M; Zande, Meike van der; Murphy, Finbarr; Sheehan, Barry; Mullins, Martin; Bouwmeester, Hans

doi:10.1080/17435390.2016.1278481

Cited by 66 publications

(62 citation statements)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to make use of the growing body of literature on ENM toxicity, a number of recent studies have proposed meta‐analysis approaches for systematic analyses of the body of evidence (Supporting Information). These studies have provided insight regarding the correlation of ENMs physicochemical properties and experimental conditions (e.g., exposure concentration and time, cell lines) with observed bioactivity . The published literature, however, is heterogeneous consisting of both quantitative and qualitative data.…”

Section: Introductionmentioning

confidence: 99%

Bayesian Network Resource for Meta‐Analysis: Cellular Toxicity of Quantum Dots

Bilal

Liu

et al. 2019

Small

View full text Add to dashboard Cite

A web-based resource for meta-analysis of nanomaterials toxicity is developed whereby the utility of Bayesian networks (BNs) is illustrated for exploring the cellular toxicity of Cd-containing quantum dots (QDs). BN models are developed based on a dataset compiled from 517 publications comprising 3028 cell viability data samples and 837 IC 50 values. BN QD toxicity (BN-QDTox) models are developed using both continuous (i.e., numerical) and categorical attributes. Using these models, the most relevant attributes identified for correlating IC 50 are: QD diameter, exposure time, surface ligand, shell, assay type, surface modification, and surface charge, with the addition of QD concentration for the cell viability analysis. Data exploration via BN models further enables identification of possible association rules for QDs cellular toxicity. The BN models as web-based applications can be used for rapid intelligent query of the available body of evidence for a given nanomaterial and can be readily updated as the body of knowledge expands. Toxicity Models www.advancedsciencenews.com

show abstract

Section: Introductionmentioning

confidence: 99%

Bayesian Network Resource for Meta‐Analysis: Cellular Toxicity of Quantum Dots

Bilal

Liu

et al. 2019

Small

View full text Add to dashboard Cite

show abstract

“…Detailed subgrouping, especially in the group of active nanomaterials should be amended to this concept, and should be related to mechanisms of action, ideally integrated into the adverse outcome pathway concept (Ankley et al 2010;Becker et al 2015;Riebeling et al 2016). A Bayesian network approach toward this concept has recently been described (Marvin et al 2017).…”

Section: Introductionmentioning

confidence: 99%

Decision tree models to classify nanomaterials according to theDF4nanoGroupingscheme

et al. 2017

Self Cite

View full text Add to dashboard Cite

To keep pace with its rapid development an efficient approach for the risk assessment of nanomaterials is needed. Grouping concepts as developed for chemicals are now being explored for its applicability to nanomaterials. One of the recently proposed grouping systems is DF4nanoGrouping scheme. In this study, we have developed three structure-activity relationship classification tree models to be used for supporting this system by identifying structural features of nanomaterials mainly responsible for the surface activity. We used data from 19 nanomaterials that were synthesized and characterized extensively in previous studies. Subsets of these materials have been used in other studies (short-term inhalation, protein carbonylation, and intrinsic oxidative potential), resulting in a unique data set for modeling. Out of a large set of 285 possible descriptors, we have demonstrated that only three descriptors (size, specific surface area, and the quantum-mechanical calculated property 'lowest unoccupied molecular orbital') need to be used to predict the endpoints investigated. The maximum number of descriptors that were finally selected by the classification trees (CT) was very low-one for intrinsic oxidative potential, two for protein carbonylation, and three for NOAEC. This suggests that the models were wellconstructed and not over-fitted. The outcome of various statistical measures and the applicability domains of our models further indicate their robustness. Therefore, we conclude that CT can be a useful tool within the DF4nanoGrouping scheme that has been proposed before. ARTICLE HISTORY

show abstract

“…In large and complex BNs (as in this case), experts cannot oversee such complexity. In a recent publication on the development of a BN to predict the hazard potential of nanoparticles this was very clear: the linkages proposed by the experts were often not supported by data (Marvin et al, 2017). In such situations, it is preferred to build the network model using data and machine learning.…”

Section: Disclaimermentioning

confidence: 99%

“…The validation of the developed BN was conducted with 20% (25163 cases) of the data randomly selected from the total dataset (Marvin et al, 2016a;Marvin et al, 2017). The validation cases were not used in the development of the BN.…”

Section: Step 3 Bn Model Validationmentioning

confidence: 99%

“…One should be aware that the methodology applied in this study results in a network based on statistical relationships only, and no other rationale for the linkages between the nodes. It is also possible to determine the linkages between the nodes and their directions by expert consultation (Marvin et al, 2017). However, such approach is less favourable in situations when complex interactions may occur (as in this situation) and in situations when sufficient data is available to allow machine learning to optimise the model.…”

Section: Bayesian Networkmentioning

confidence: 99%

See 1 more Smart Citation

Applicability of a food chain analysis on aquaculture of Atlantic salmon to identify and monitor vulnerabilities and drivers of change for the identification of emerging risks

Marvin

Bouzembrak

Asselt

et al. 2019

EFSA Supporting Publications

Self Cite

View full text Add to dashboard Cite

The objective of Aquarius project was to test the applicability of a food chain analysis on Atlantic salmon farmed in Norway to identify and monitor vulnerabilities and drivers of change for the identification of emerging risks. To this end, a comprehensive literature review was conducted complemented with expert views collected by on‐line questionnaires and in‐depth interviews. This review provided an overview of the various stages in the supply chain from farm to consumer and the identification of i) existing and emerging human and animal health hazards, ii) vulnerabilities, iii) control measures, iv) drivers of change, and v) related indicators. Next step in the study was to i) complement the list of vulnerabilities followed by a prioritisation using focus group discussions and Failure Mode and Effect Analysis (FMEA) method and ii) link drivers of change acting upon the (prioritised) vulnerabilities and identification of associated indicators and data sources by means of a Delphi method. Based on the experiences obtained with the different methodologies during this project, the Aquarius team considers literature review, and a direct face‐to‐face interaction with experts (in‐depth interviews, focus group discussions and FMEA in a workshop setting) as being most effective regarding output and effort. The comprehensive information collected formed the basis for the methodology to baseline key indicators of the selected vulnerabilities. The methodology proposed comprised a Bayesian Network (BN) that links data sources for indicators and drivers of change to the prioritised vulnerabilities and was demonstrated for salmon health. A Decision Support System (DSS) with such BN integrated was designed. The current BN and DSS could flag an increased exposure of known hazards that may occur due to changes in indicators but new hazards are not identified. The proposed methodology could be used instead to show trends in vulnerabilities, indicators, increased exposure of known hazards and for scenario studies.

show abstract

Application of Bayesian networks for hazard ranking of nanomaterials to support human health risk assessment

Cited by 66 publications

References 59 publications

Bayesian Network Resource for Meta‐Analysis: Cellular Toxicity of Quantum Dots

Bayesian Network Resource for Meta‐Analysis: Cellular Toxicity of Quantum Dots

Decision tree models to classify nanomaterials according to theDF4nanoGroupingscheme

Applicability of a food chain analysis on aquaculture of Atlantic salmon to identify and monitor vulnerabilities and drivers of change for the identification of emerging risks

Contact Info

Product

Resources

About