Understanding the relationships between the physicochemical properties of engineered nanomaterials and their toxicity is critical for environmental and health risk analysis. This task is severely confounded by material diversity, heterogeneity of published data, and limited sampling within individual studies. A meta-analytical approach for extracting and analyzing pertinent knowledge from published studies was previously developed to focus on the cellular toxicity of Cd-containing semiconductor quantum dots (QDs). From literature mining of 307 publications, we obtained 1,741 cell viability-related data samples, each with 24 qualitative and quantitative attributes describing the material properties and experimental conditions. Developing and applying random forest regression models to the data showed that toxicity was closely correlated with several QD surface properties including shell material, ligand and surface modifications, diameter, assay type and, of course, cellular exposure time. Following up from that study, a newly developed resource for meta-analysis of nanomaterial toxicity utilizing Bayesian Networks (BNs) was also tested for exploring the cellular toxicity of Cd-containing QDs. Here, the previously extracted dataset was significantly expanded to now include 517 publications comprising 3,028 cell viability data samples and 837 IC50 values. BN-QD toxicity models were developed using both continuous or numerical and categorical attributes. Using these models, the most relevant attributes identified for correlating IC50 to cellular toxicity were similar to that of the previous study and included: QD diameter, exposure time, surface ligand, shell, assay type, surface modification, and surface charge, and QD concentration for the cell viability analysis. Intense data exploration with the BN models further enabled identification of possible multi-parametric association rules for QDs cellular toxicity. This overall approach of integrating quantitative and categorical data provides for interrogating wide-ranging toxicity data in the literature data suggests that meta-analysis can help develop methods for predicting the toxicity of engineered nanomaterials. Moreover, the available body of evidence for a given nanomaterial can be readily updated as the body of knowledge expands with more studies.

Utilizing meta-analysis to understand the cellular toxicity of quantum dots

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.