Research

Why model toxicology ?

For years, the use of chemical products in our everyday life has steadily increased and the toxicity of these substances is becoming an area of great concern within society. Such concern has prompted the strengthening of regulations regarding the production and use of chemical substances worldwide. As a consequence, chemical manufacturers must now conduct extensive toxicity studies to demonstrate the innocuousness of their products, sky-rocketing the development cost of such products. This context provides ground for the use of computational methods to model toxicity.

Quantitative models are data-expensive

From toxicity prediction to toxicity pathways exploration

So far, most of the modelling approaches in toxicology are quantitative. They aim at either inferring the toxic threshold of a chemical substance or confirming its specific pathway of toxicity, namely, the detailed mechanism by which a substance disrupts the physiological equilibrium of an organism. These objectives require a lot of biological data, which can be restrictive given the current acquisition cost of such data. An alternative approach consists in shifting the focus from toxic thresholds to toxicity pathways. Indeed, describing these pathways in a qualitative manner would allow to focus only on equilibrium changes and would therefore require comparatively less biological data.

Using qualitative models in toxicology

Several generic formalisms have already been developed to qualitatively model biological processes [1, 2, 3, 4, 5]. These formalisms use formal methods to reason about these standard processes. However, expressing toxicology problems in manageable terms for the formalism is frequently troublesome. For instance, these formalisms describe chemical reactions and only depict equilibria as indirect results of competing rule kinetics. Yet, toxicity pathways are sequences of equilibrium changes. Keeping equilibria implicit while building a toxicological model can thus hinder the identification of possible toxicity pathways.

To solve these limitations, my research focuses on the development of a domain-oriented formalism directly describing qualitative equilibrium changes. First, a rule-based language allows to express the different equilibrium changes present in a biological system. Then, the chaining of rules can be corseted thanks to constraints expressed in an extended temporal logic. These constraints are usually based on toxicological observations regarding specific conditions of the system. Finally, automated reasoning tools can be used on the resulting system dynamics to detect possible toxicity pathways, providing useful insights to improve the experimental strategies of toxicity studies. [More details soon !]