Periodic Reporting for period 1 - INSIGHT (Integrated Models for the Development and Assessment of High Impact Chemicals and Materials)
Reporting period: 2024-01-01 to 2025-06-30
Europe’s Green Deal and the Chemicals Strategy for Sustainability call for chemicals and materials that are safe and sustainable by design (SSbD). Today, decisions about new substances often rely on fragmented evidence: hazard studies, exposure estimates, life-cycle assessments and socio-economic analyses are performed independently from each other, with different assumptions and formats. This slows innovation, increases costs, and makes it difficult for regulators and industry to compare options transparently.
The INSIGHT project addresses this challenge by developing and testing an integrated, mechanistic framework that connects evidence from molecules to market. INSIGHT links computational and experimental information into Impact Outcome Pathways (IOPs), which are chains of cause-and-effect that start from molecular initiating events and exposure, and lead to health, environmental, social and economic outcomes. Around these pathways, INSIGHT is assembling:
- a data graph (curated, FAIR datasets with open programmatic access),
- a model graph (documented models with clear inputs/outputs and versioning),
- pipelines that combine models and data to answer concrete questions, and
- decision maps and a user interface that guides non-experts through transparent, reproducible assessments.
Four case studies anchor the development of the INSIGHT framework to real-world needs (advanced batteries/construction materials, aerospace polymers, tyre fillers and antimicrobial coatings). Social sciences inform the design of the socio-economic pathways, stakeholder perspectives and the analysis of trade-offs, ensuring the framework supports decisions that are technically sound and societally acceptable.
By the end of the project, the INSIGHT consortium expects to deliver a pilot-ready platform that reduces time-to-decision, improves reproducibility, and supports regulatory and industrial uptake of SSbD across Europe.
The INSIGHT project addresses this challenge by developing and testing an integrated, mechanistic framework that connects evidence from molecules to market. INSIGHT links computational and experimental information into Impact Outcome Pathways (IOPs), which are chains of cause-and-effect that start from molecular initiating events and exposure, and lead to health, environmental, social and economic outcomes. Around these pathways, INSIGHT is assembling:
- a data graph (curated, FAIR datasets with open programmatic access),
- a model graph (documented models with clear inputs/outputs and versioning),
- pipelines that combine models and data to answer concrete questions, and
- decision maps and a user interface that guides non-experts through transparent, reproducible assessments.
Four case studies anchor the development of the INSIGHT framework to real-world needs (advanced batteries/construction materials, aerospace polymers, tyre fillers and antimicrobial coatings). Social sciences inform the design of the socio-economic pathways, stakeholder perspectives and the analysis of trade-offs, ensuring the framework supports decisions that are technically sound and societally acceptable.
By the end of the project, the INSIGHT consortium expects to deliver a pilot-ready platform that reduces time-to-decision, improves reproducibility, and supports regulatory and industrial uptake of SSbD across Europe.
- Pathway framework: Building on Adverse Outcome Pathways, a modular Impact Outcome Pathway (IOP) framework was extended beyond the health outcomes of chemicals/materials, towards environmental, social and economic dimensions. Literature-based pathways from life-cycle and socio-economic assessment have been translated into this structure, forming a common backbone for cross-dimension integration.
- Models and pipelines: a total of 93 models/tools relevant to health, environmental and socio-economic assessment have been catalogued; 36 already expose APIs or container images for direct integration. The first end-to-end mechanistic pipeline was initiated for per- and polyfluoroalkyl substances (PFAS), chaining dose-response modelling (benchmark dose), mechanistic pattern recognition and physiologically-based pharmacokinetics to connect in vitro evidence with plausible human exposure and effect levels.
- Data and FAIRification: A Data Management Plan was delivered and the FAIRification workflow is being implemented. Twelve priority datasets (e.g. toxicology, ecotoxicology and exposure sources) are being harmonised with ontologies and made accessible through a public API. Three curated datasets associated with the case studies have been cleaned and annotated for direct use in IOPs.
- Integration architecture: The model graph and data graph are being linked through explicit, machine-readable input–output relations, enabling automated pipeline assembly and provenance tracking. Utility nodes (for consensus and integration) are being defined to support robust, reusable computations.
-Tools and usability: A first module of the graphical interface (an assisted MODA documentation tool for model reporting) has been prototyped to ease standard-compliant documentation and reuse
- Models and pipelines: a total of 93 models/tools relevant to health, environmental and socio-economic assessment have been catalogued; 36 already expose APIs or container images for direct integration. The first end-to-end mechanistic pipeline was initiated for per- and polyfluoroalkyl substances (PFAS), chaining dose-response modelling (benchmark dose), mechanistic pattern recognition and physiologically-based pharmacokinetics to connect in vitro evidence with plausible human exposure and effect levels.
- Data and FAIRification: A Data Management Plan was delivered and the FAIRification workflow is being implemented. Twelve priority datasets (e.g. toxicology, ecotoxicology and exposure sources) are being harmonised with ontologies and made accessible through a public API. Three curated datasets associated with the case studies have been cleaned and annotated for direct use in IOPs.
- Integration architecture: The model graph and data graph are being linked through explicit, machine-readable input–output relations, enabling automated pipeline assembly and provenance tracking. Utility nodes (for consensus and integration) are being defined to support robust, reusable computations.
-Tools and usability: A first module of the graphical interface (an assisted MODA documentation tool for model reporting) has been prototyped to ease standard-compliant documentation and reuse
INSIGHT advances the state of the art by introducing a single mechanistic language, Impact Outcome Pathways (IOPs), that links health, environmental, social and economic evidence and makes trade-offs explicit. It builds a graph-based computational ecosystem in which FAIR, versioned datasets and models are plug-and-play, and it provides decision maps that guide non-experts through transparent, reproducible workflows. Early end-to-end pipelines already connect in vitro effects to exposure and population-level endpoints, paving the way for next-generation risk assessment and SSbD screening.
These advances are expected to shorten and lower the cost of assessments by i) reusing curated data and preconfigured pipelines, ii) reducing duplication and animal testing by relying more on mechanistic and in silico evidence, and iii) clarifying benefits versus trade-offs, such as performance, toxicity and climate impacts, to support responsible innovation.
To enable broader usability and impact, the project will complete and harden the user interface and APIs, benchmark performance and reproducibility on the case studies, continue FAIRification and standardisation of data and models with persistent identifiers and machine-readable reporting aligned to widely used templates, and provide validation and documentation that meet regulatory expectations on provenance, uncertainty and applicability domains. Long-term sustainability will be supported through robust hosting and maintenance, training for regulators and industry, and clear licensing that enables open use and, where appropriate, commercial services.
These advances are expected to shorten and lower the cost of assessments by i) reusing curated data and preconfigured pipelines, ii) reducing duplication and animal testing by relying more on mechanistic and in silico evidence, and iii) clarifying benefits versus trade-offs, such as performance, toxicity and climate impacts, to support responsible innovation.
To enable broader usability and impact, the project will complete and harden the user interface and APIs, benchmark performance and reproducibility on the case studies, continue FAIRification and standardisation of data and models with persistent identifiers and machine-readable reporting aligned to widely used templates, and provide validation and documentation that meet regulatory expectations on provenance, uncertainty and applicability domains. Long-term sustainability will be supported through robust hosting and maintenance, training for regulators and industry, and clear licensing that enables open use and, where appropriate, commercial services.