Assessment of Progress - Research Progress on Environmental Health and Safety Aspects of Engineered Nanomaterials - page 26

On the basis of the still unmet need for more data-integration mechanisms, the committee has
characterized this indicator as yellow.
 Advancement of systems for sharing the results of research and fostering development of
predictive models of nanomaterial behaviors
In its first report, the committee identified the need to develop predictive models for ENM
behaviors and risk. However, the development of models cannot occur in isolation from data generation.
Coordination is needed in the short term to ensure that experimental, modeling, and informatics efforts
contribute to a coordinated, functional infrastructure. There is a need to collect, store, archive, and share
data related to assessing the potential effects of ENMs (as described in the previous section) so that these
data can be used to develop predictive models of ENM behavior. The goals of advancing systems for
sharing and developing models of behavior are intimately related in that the models and data structures
are both influenced by the specific questions related to exposure to ENMs and the resulting effects that
need to be addressed. Therefore, the needs for models and infrastructure to support the models are
assessed together.
There has been some progress in development of models to predict nanomaterial exposures and
toxicity (Gottschalk et al. 2011; Nel et al. 2013). Several government agencies have instituted specific
programs to develop and test different models to assess ENM behavior (for example, fate in the
environment, releases from consumer products, plant uptake, and occupational exposure), including EPA,
NIST, FDA, DOD, the US Department of Agriculture, and NIOSH (NSET 2012a). Efforts are also in
place to develop computational models for toxicity (for example, EPA's ToxCast program). Finally, there
has been progress towards the development of empirical predictive models as opposed to fully
mechanistic models for behavior (Hou et al. 2013; Westerhoff and Nowack 2013). These models rely on
empirical correlations (for example, partition coefficients) rather than complete mechanisms. The models
can be developed in less time than fully mechanistic models, and can predict approximate behaviors (for
example, in a wastewater treatment plant) and may be used to support regulatory decisions.
The committee classifies progress in this category as yellow because, despite the development
and use of the models in the nanotechnology-EHS community, there is not yet a central repository for
sharing the models (although NanoHub may be appropriate), and many needed models have not yet been
developed, such as models to predict the structure of ENM surfaces in various environments. Most
important, there is a paucity of data for calibrating and validating models that have been developed; for
example, there are very few data on ENM concentrations and speciation in environmental and biologic
media that can be used to calibrate fate and transport or biodistribution models. The absence of metadata
and validation data for most models hampers their broad acceptance and use because they are not deemed
reliable and accurate.
Some progress is being made in the collection, storage, and archiving of ENM physical and
chemical properties. For example, the Nanomaterials Registry (NR) has been developed by the Research
Triangle Institute with funding from NIH (Nanomaterialregistry 2013). The NR will provide a curated
repository of ENM information (for example, ENM properties) from a wide array of studies that used the
materials. The repository would allow researchers to compare model results for behaviors and effects of
ENMs by using data on the nanomaterials stored in the NR. Incorporation of information on biologic and
environmental interactions in the NR is also being considered. Other databases are being created (for
example, the Nano-Bio Interactions Knowledgebase) with similar aims: to capture and store information
about nanomaterial properties and behaviors that allow development of structure-activity relationships
and other scientific synthesis using large datasets. Finally, the new standard data format, ISA-TAB-Nano,
for sharing results obtained with analytic methods for characterization of nanomaterial properties and
effects has recently been published (Thomas et al. 2013).
The committee classifies progress in this category as yellow because despite initial efforts and
models developed, the models and data are not yet widely available and there is no agreement about the
appropriate architecture for the databases, no agreement about ontology (although it is being developed