Our SERS Computational Science capability focuses on delivering human and environmental solutions that enable safety assessments without animal testing and accelerate the transition towards NGRA. We combine advanced data analytics, mechanistic modelling, and expert-led, AI-augmented workflows to support the safety of our ingredients and products. By integrating diverse data streams, from molecular properties and bioactivity signatures to in vitro and exposure-led insights, we build transparent, robust and scientifically credible tools and approaches to support our safety assessments.
Computational approaches play a crucial role in understanding exposure. Within SERS, physiologically based kinetic (PBK) models are developed to translate external exposure into internal, tissue-specific concentrations, thereby enabling human- or environmental species-relevant interpretation of in vitro bioactivity data (see additional details in our Exposure Science page). Alongside this, we apply environmentally relevant models, including toxicokinetic, bioaccumulation and environmental fate models, to understand how substances move through ecosystems. These computational tools allow us to generate realistic exposure estimates that inform the weight of evidence for both human and environmental safety (see additional details in our Systemic Safety and Environmental Safety pages).
We use a wide range of in silico hazard prediction tools and quantitative new approach methodologies (NAMs) to identify potential risks early and prioritise targeted testing strategies. In addition to applying best practice methods for analysing in vitro data, we also develop novel approaches to characterising bioactivity, using Bayesian concentration-response modelling. Mechanistic insights are further informed through bioinformatic and cheminformatic analyses, helping us understand how and why substances interact with biological systems.
In the field of Skin Allergy Safety, we developed the Skin Allergy Risk Assessment (SARA) model, which integrates real‑world human skin exposure with non-animal NAM data and other human relevant lines of evidence. This tool enables us to predict the skin sensitiser potency of an ingredient and derive a scientifically robust PoD for assessing risk at consumer exposure levels. This work has also underpinned the development of the SARA-ICE Defined Approach, which is currently accepted in OECD GL No. 497.
To support cross‑species extrapolation of biological activity induced by a test substance, we developed G2P‑SCAN (Gene‑to‑Pathways Species Conservation Analysis), a computational pipeline that maps human genes and pathways across selected model species using orthology and functional family information. G2P‑SCAN provides biologically meaningful insights to support animal-free Environmental Safety decision‑making and has been included within OECD Integrated Approaches to Testing and Assessment (IATA) for cross‑species extrapolation in environmental safety (OECD Series on Testing and Assessment, No. 415 (PDF 2.8 MB)).
