S5: Predictive Toxicology and Safety Assessment
Knowledge-based Integration supporting Predictive Toxicology and Safety Assessment
To make a good decision we need to bring both expertise and relevant information together to form the basis for a structured well-informed discussion leading to best judgement based on available evidence and opinions formed on it. Such a knowledge integration is required in many areas of toxicology and safety assessment based on scientific knowledge generated by a growing number of alternative testing research methods and initiatives. Integration may include evidence from in vitro or in silico methods, biology or chemistry, science and engineering, human health or environment-oriented, and requires both effective organisation of knowledge and communications to reach common understandings.
The requirements, context and format for applications may vary e.g., asking and answering scientific questions, carrying out a risk assessment on products, or performing a regulatory decision or submission. All applications however require a sound reproducible scientific basis and the use of good practices in characterising experiments, organising data and describing concepts in our ontology and knowledge framework.
There is a growing opportunity to develop knowledge integration based on combining emerging concepts and frameworks e.g., OpenTox for data integration and resource interoperability, adverse outcome pathways for mapping data to events, Risk21 for a combined analysis of exposure and hazard, and weight of evidence and read across methods for combining evidence from chemical or biological mechanistic categories. We also need community frameworks to bring different disciplines together for fruitful interactions and discussions.
In this presentation I will review recent developments related to OpenTox including recent work on knowledge resource interoperability and workflows supporting data access and integration, integrated testing strategies, systems biology modelling and risk assessment against AOPs.