D5.4 - First predictive nanoQSAR models integrated into NanoCommons KnowledgeBase

This Deliverable report (D5.4) presents the services that have been developed and are available through the NanoCommons infrastructure for generating and validating nano-specific quantitative Structure-Activity Relationships (nanoQSAR) models and applying the models for predicting nanomaterial (NM) end-points for new materials that have not been tested experimentally. We offer two levels of modelling services: (i) Algorithm as a Service (AaaS) which provides the technical tools to model developers for creating nanoQSAR models and deploying them in the NanoCommons Knowledge Base, or (ii) Model as a Service (MaaS) which provides ready-to-use web implementations of nanoQSAR models, that can be used by the community to validate the models or calculate end-point predictions for other NMs as long as these are within the Domain of Applicability (DoA) of the model.