RISK CDT - Rational Decision-Making based on Predictive Toxicology

PLEASE APPLY ONLINE TO THE SCHOOL OF ENGINEERING, PROVIDING THE PROJECT TITLE, NAME OF PRIMARY SUPERVISOR AND SELECT THE PROGRAMME CODE "EGPR" (PHD - SCHOOL OF ENGINEERING)

This is a project within the multi-disciplinary EPSRC and ESRC Centre for Doctoral Training (CDT) on Quantification and Management of Risk & Uncertainty in Complex Systems & Environments, within the Institute for Risk and Uncertainty. The studentship is granted for 4 years and includes, in the first year, a Master in Decision Making under Risk & Uncertainty. The project includes extensive collaboration with prime industry to build an optimal basis for employability.

The sophistication of predictive toxicology and pharmacology approaches is increasing as is the demand for their use to address critical needs in safety assessment. This presents challenges both for developers and those making decisions based on the predictions since often these two communities do not share the same level of confidence in an approach put forward for a particular purpose. Recent legislative changes have shifted the emphasis towards computational biology in toxicology and pharmacology for information required for safety and risk assessments; however, in many cases there is a sparsity of empirical data to support the development and validation or confirmation of computational models. Recent work has established a framework for establishing the credibility of simulations in computational biology, i.e. the willingness to make decisions based on the outcomes of simulations. In current work, a framework to establish the scientific credibility of predictive toxicology approaches through a process of social epistemology has been proposed. An essential tool within the framework is a ’credibility’ matrix which facilitates a collaborative means of determining the extent to which a predictive approach is underpinned by knowledge and understanding of the biological system and by empirical data for real-world referentiality. A set of seven credibility factors have been proposed, based on well-established philosophical foundations, that serve to guide the systematic and objective assessment of predictive approaches and to guide efficient and effective validation strategies to strengthen the evidence base. These factors are closely related to, but not identical to, factors considered in confirming or validating models in other fields of computational science and engineering.

In this project, it is proposed to extend this current work by considering the uncertainty associated with these credibility factors, including the underlying assumptions upon which a predictive approach is established. It is anticipated that this will be tackled by reviewing uncertainty quantification methods used in other fields and exploring their applicability in predictive toxicology and pharmacology, as well exploring innovative approaches to uncertainty quantification, including developing an uncertainty budget associated with a predictive approach. While uncertainty quantification is an important aspect of establishing credibility for predictive approaches, the willingness of decision-makers to accept predictive approaches is also based on tacit knowledge and understanding of the predictive approach and its underlying assumptions and theoretical ancestry. This tacit information is often very familiar to the researchers and developers of predictive approaches but relatively unfamiliar to decision-makers and other stakeholders. Hence, the project will also explore the current mechanisms of social networking and epistemology at work in the toxicology and pharmacology regulatory environment, using methods developed in other fields such as engineering product design; and, consider potential enhancements of the mechanisms in the regulatory environment that will increase credibility and application of predictive approaches. Finally, these two stands of uncertainty quantification and social epistemology will be brought together to develop a decision-making methodology that enhances the regulatory process in toxicology and pharmacology in terms of robustness and reliability as well as increasing public understanding and trust.

The research project is a collaboration between the School of Engineering and the Institute for Risk and Uncertainty at the University of Liverpool and the Directorate for Health, Consumers and Reference Materials at the European Commission Joint Research Centre in Ispra, Italy. It is anticipated that the outcomes of the research project will have applications beyond toxicology and pharmacology in other fields of science and engineering where the predictions of computational models are used to inform decisions based on sparse information.