378
E. Benfenati et al.
Similarly, these criteria (reliability and consistency) are adopted within JANUS.
The tools within VEGAHUB are transparent. The final decision is taken by the
user, while VEGAHUB offers support representing the different lines of evidence.
Thus, the user may opt for a conservative result, or prefer a realistic value.
We notice that different tools relate to different purposes. Users may be interested
in the specific evaluation of a single chemical and apply a range of tools for this
need. Alternatively, the user may be interested in screening, discriminating families
of chemicals of concern from the safe one. In other cases, the user may need prioritization criteria. While groups of chemicals screened together may be treated together
(PBT or not, CMR or not), for prioritization a list is necessary, because substances
have to be ordered. Indeed, the user may have resources to separately assess a limited
number of compounds, and it is necessary to provide a way to sort this number out of
the group of CMR chemicals, for instance. For this reason, within JANUS we developed models that provide continuous values for carcinogenicity and reprotoxicity.
For instance, carcinogenic compounds can be ordered according to their potency
value.
18.5 Conclusion and Future Directions
Within VEGAHUB we provide a series of tools with the aim to better evaluate
individual chemicals, or families of substances. The individual tools, QSAR, readacross, and grouping, are related to the methodological aspects. We tried to move
toward results of the models, which provide details, interpretations, related values,
occurrence of SA, graphics with associations and trend, etc. beyond the technicalities
of the individual models. Within VEGAHUB multiple tools are connected and used
simultaneously in order to assist the user.
We have already planned extensions of VEGAHUB. A major improvement will
be achieved by joining VEGAHUB with MERLIN-Expo (https://merlin-expo.eu/),
within the LIFE VERMEER project (https://www.life-vermeer.eu/) funded by the
EC. In this way, the user will have combined tools for the hazard and for the exposure, and thus, the evaluation of risk assessment will be facilitated. Furthermore,
VERMEER will identify safer substitutes of risky chemicals, and in this, it will be
aligned with the principles of the REACH legislation.
The toDIVINE project funded by the German UBA will further integrate readacross (based on substances registered for REACH) with QSAR predictions.
The LIFE CONCERT REACH project, funded by the EC, will establish a network
between VEGAHUB, the Danish QSAR database, and AMBIT (http://cefic-lri.org/
toolbox/ambit/), to fully exploit data registered within REACH. All endpoints will be
scrutinized, in an effort to develop QSAR models for the tens of endpoints addressed
within REACH.
The Optitox project, funded by EFSA, will derive QSAR models from the EFSA
OpenFoodTox database by leveraging information about the toxicokinetics.
E. Benfenati et al.
Similarly, these criteria (reliability and consistency) are adopted within JANUS.
The tools within VEGAHUB are transparent. The final decision is taken by the
user, while VEGAHUB offers support representing the different lines of evidence.
Thus, the user may opt for a conservative result, or prefer a realistic value.
We notice that different tools relate to different purposes. Users may be interested
in the specific evaluation of a single chemical and apply a range of tools for this
need. Alternatively, the user may be interested in screening, discriminating families
of chemicals of concern from the safe one. In other cases, the user may need prioritization criteria. While groups of chemicals screened together may be treated together
(PBT or not, CMR or not), for prioritization a list is necessary, because substances
have to be ordered. Indeed, the user may have resources to separately assess a limited
number of compounds, and it is necessary to provide a way to sort this number out of
the group of CMR chemicals, for instance. For this reason, within JANUS we developed models that provide continuous values for carcinogenicity and reprotoxicity.
For instance, carcinogenic compounds can be ordered according to their potency
value.
18.5 Conclusion and Future Directions
Within VEGAHUB we provide a series of tools with the aim to better evaluate
individual chemicals, or families of substances. The individual tools, QSAR, readacross, and grouping, are related to the methodological aspects. We tried to move
toward results of the models, which provide details, interpretations, related values,
occurrence of SA, graphics with associations and trend, etc. beyond the technicalities
of the individual models. Within VEGAHUB multiple tools are connected and used
simultaneously in order to assist the user.
We have already planned extensions of VEGAHUB. A major improvement will
be achieved by joining VEGAHUB with MERLIN-Expo (https://merlin-expo.eu/),
within the LIFE VERMEER project (https://www.life-vermeer.eu/) funded by the
EC. In this way, the user will have combined tools for the hazard and for the exposure, and thus, the evaluation of risk assessment will be facilitated. Furthermore,
VERMEER will identify safer substitutes of risky chemicals, and in this, it will be
aligned with the principles of the REACH legislation.
The toDIVINE project funded by the German UBA will further integrate readacross (based on substances registered for REACH) with QSAR predictions.
The LIFE CONCERT REACH project, funded by the EC, will establish a network
between VEGAHUB, the Danish QSAR database, and AMBIT (http://cefic-lri.org/
toolbox/ambit/), to fully exploit data registered within REACH. All endpoints will be
scrutinized, in an effort to develop QSAR models for the tens of endpoints addressed
within REACH.
The Optitox project, funded by EFSA, will derive QSAR models from the EFSA
OpenFoodTox database by leveraging information about the toxicokinetics.