Addressing a bottle neck for regulation of nanomaterials: quantitative read-across (Nano-QRA) algorithm for cases when only limited data is available

Abstract

The development of in silicomethods that support human health and environmental risk assessment of engineered nanomaterials is nowadays of high interest, because the application of those methods enables to fill the existing experimental data gaps.

Abstract

The number and variety of engineered nanoparticles have been growing exponentially. Since the experimental evaluation of nanoparticles causing public health concerns is expensive and time consuming, efficient computational tools are amongst the most suitable approaches to identifying potential negative impacts, to the human health and the environment, of new nanomaterials before their production. However, developing computational models complimentary to experiments is impossible without incorporating consistent and high quality experimental data. Although there are limited available data in the literature, one may apply read-across techniques that seem to be an attractive and pragmatic alternative way of predicting missing physico-chemical or toxicological data. Unfortunately, the existing methods of read-across are strongly dependent on the expert's knowledge. In consequence, the results of estimations may vary dependently on personal experience of expert conducting the study and as such cannot guarantee the reproducibility of their results. Therefore, it is essential to develop novel read-across algorithm(s) that will provide reliable predictions of the missing data without the need to for additional experiments. We proposed a novel quantitative read-across approach for nanomaterials (Nano-QRA) that addresses and overcomes a basic limitation of existing methods. It is based on: one-point-slope, two-point formula, or the equation of a plane passing through three points. The proposed Nano-QRA approach is a simple and effective algorithm for filling data gaps in quantitative manner providing reliable predictions of the missing data.