GPC Research & Development Programme 2026
Objectives, research streams, collaborations and results, 2019–2028
Global Product Compliance (Europe) AB – internal document, October 2026
1. Purpose and overall aim
GPC’s research and development programme exists to make chemical regulation and compliance more scientific, more efficient and less dependent on new testing, especially animal testing. It turns GPC’s practical experience as a regulatory partner and Lead Registrant into research questions, answers them with academic and industry partners, and builds the results into tools and services.
The programme is organised around five research objectives:
No. | Objective | Expected outcome |
1 | Regulatory science for compliance excellence and efficiency | Systematic knowledge of how regulations are built and depend on each other, turned into real-time, portfolio-specific compliance guidance (Intrims, RegAlert) |
2 | Avoidance of unnecessary testing | Criteria (DIM), read-across, QSAR and machine-learning models (Jadavpur University), and a methodology that show when new studies, especially animal studies, can be scientifically avoided |
3 | Group-based chemical safety assessment | Approaches for assessing groups of chemicals, with a reference set of about 150 chemical safety profiles as case examples |
4 | Data exchange and data mining | Platforms that find existing data, interpret it against regulatory requirements and make it available for reuse (Datagap, Chemsortia) |
5 | Cost-efficient and realistic biodegradation testing | Methods that cut the cost and time of degradation kinetics experiments by at least 50–60% and simulate real, complex effluents |
The objectives build on each other: regulatory science (1) defines what data are needed; testing avoidance (2) and group assessment (3) show how needs can be met without new studies; the data platforms (4) find and share the data that already exist; and better biodegradation methods (5) make the studies that are still needed cheaper and more relevant.
2. GPC R&D at a glance
Item | Summary |
Base | Ideon Science Park, Lund (Beta 5, Scheelevägen 17) |
Core R&D team | Shisher Kumra, Dr Anders Bergqvist, Dr Zsuzsanna Szepesi, Dr Vaibhav Diwan, Mirach Pelster |
Main partners | DTU (Denmark), IIT Bombay, Lund University (incl. LTH), Jadavpur University, the Animal-Free Safety Assessment (AFSA) Collaboration, KTH, RISE, S&P Global |
Funding | GPC’s own resources; co-financing of university projects (DTU); Vinnova-funded project with GPC in kind |
Publications | Including four papers with Jadavpur University on read-across, QSAR and machine-learning models; peer-reviewed papers in Regulatory Toxicology and Pharmacology, Chemical Research in Toxicology, SAR and QSAR in Environmental Research and Aquatic Toxicology; further manuscripts and a book chapter in progress |
Intellectual property | Indian patent application 202521067103 (Intrims technology), filed July 2025, published September 2026 |
Products from research | Intrims (launch 2027), RegAlert, Datagap (launch December 2026), Chemsortia (launch May 2027) |
Training | More than 20 interns and thesis students since 2019, mainly from Lund University |
3. People and roles
Personne | Role | Objectives |
Shisher Kumra | Leads the R&D programme and the regulatory science research; concept and inventor of Intrims; concept for Datagap and Chemsortia | 1, 4 (and PFAS within 1) |
Dr Anders Bergqvist | Head of toxicology (PhD Immunology, Lund University); leads the chemical safety profiles and the Handbook | 2, 3 |
Dr Zsuzsanna Szepesi | Researcher, in silico toxicology (PhD Immunology, Lund University); read-across, QSAR and machine-learning models with Jadavpur University; chemical safety profiles | 2, 3 |
Dr Vaibhav Diwan | Internship and DTU Research Manager (PhD, DTU 2019); also Business Development & Growth | 1, 4, 5 |
Mirach Pelster | Researcher and developer (chemical engineer, MSc) | 1, 4 |
Regulatory dossiers and other client deliverables are prepared by GPC’s regulatory staff, so that R&D time and client work can be kept apart.
4. Objective 1 – Regulatory science for compliance excellence and efficiency
Why it matters. Chemicals regulations worldwide differ in structure, definitions and requirements, and companies struggle to know what they must do, in which order, for each substance and market. Systematic analysis of regulations makes compliance faster, cheaper and more reliable, and helps countries design better regulation.
Research questions. How are chemicals regulations structured, and how do their requirements depend on each other? Can best practice be identified and adapted to different countries? Can a country’s chemicals inventory be predicted from its industrial structure? How can this knowledge be turned into real-time, company-specific guidance?
Activities and results
Activity | Period | Key result |
Chemicals inventory estimation model | 2022 – ongoing | Predicts a country’s inventory from its industry structure (manufacturing-, application- or consumer-intensive) and economic level; tested on three countries with 75–80% correct results; online tool in the Global Inventory Report (January 2025) |
Global regulatory framework and best practice | 2022 – 2026 | Regulations evaluated against about 18 criteria; best practice per criterion; overall model adaptable to countries; Global Chemical Regulatory Framework (November 2026) |
Survey on the impact of chemical safety regulation | 2025 | 170 industry respondents; most see a lack of regulation as harmful to business and to a level playing field (Asia Chemicals Forum 2025) |
Analysis of regulatory selection of substances for evaluation | 2024 – July 2025 | How selection for evaluation compares with substances of priority concern |
PFAS identification and regulatory implications | 2025 – March 2027 | Computational method to identify PFAS and their sub-groups and link them to legal implications; thesis, white paper with S&P Global, IPI article |
Regulatory barriers to continuous manufacturing of biologics (Vinnova, Lund University) | 2024 – 2025 | GPC as regulatory partner; manuscript submitted (2026); EUFEPS presentation |
From research to products
- Intrims (2024 – launch 2027): standardises the requirements of all applicable regulations into one set of compliance actions, groups them into about 20 bundles by interdependency (OR, AND, MIXED, SINGLE logics), sequences the bundles and applies them to a company’s own portfolio, with automatic re-evaluation. Tested with more than 40 users (2024) and more than 25 users (2025); upgraded after testing. Patent application 202521067103 (India).
- RegAlert: real-time regulatory alerts based on the same knowledge base.
Data foundation. Annual GPC reviews of chemicals regulations: Global Regulatory Guide, GHS Compliance guide, Global Inventory Report, Annual Summary, Compendium of Global Chemical Regulation, and the PFAS regulation overview.
5. Objective 2 – Avoidance of unnecessary testing
Why it matters. Many studies requested under REACH, including animal studies, are difficult, impossible or meaningless for certain substances, or could be replaced by existing data and non-animal methods.
Research questions. When is an animal test difficult, impossible or meaningless? Which read-across, QSAR and machine-learning models can reliably replace specific studies? How can a Lead Registrant decide, case by case, whether a requested study can be scientifically avoided?
Activities and results
Activity | Partner | Period | Key result |
DIM criteria for difficult, impossible and meaningless animal testing | AFSA Collaboration | Feb 2025 – Aug 2026; next phase to Q3 2027 | Criteria applied to 1,740 substances; 35 avoidable fish studies, about 5,000–10,000 animals; van Velthoven et al. (2026) |
Read-across, QSAR and machine-learning models (details below) | Jadavpur University | 2023 – 2026, continuing | 15 peer-reviewed publications from the collaboration, 4 with GPC co-authors; micronucleus model submitted (2026) |
Testing-avoidance methodology as Lead Registrant | Internal | 2022 – ongoing | Combines existing data, read-across, models and DIM criteria; study requirements reduced in about 10–15% of the substances assessed in 2025–2026 |
Earlier work: REACH last-resort requirement; difficult and coloured substances | AFSA; DTU | 2019 – 2024 | Macmillan et al. (2024); DTU project 2019–2021 |
In silico and read-across methods with Jadavpur University
Together with the Drug Theoretics and Cheminformatics Laboratory at Jadavpur University (Prof. Kunal Roy), GPC develops computational methods that can replace or reduce animal studies: read-across, QSAR, machine-learning models and combined approaches such as q-RASAR (quantitative read-across structure-activity relationships). GPC formulates the research questions from its experience of animal studies and of gaps in existing models, compiles training data from its own and public sources, defines the regulatory endpoints and applicability needs, and helps adapt the final models. The collaboration, running in two projects since 2023, has produced 15 peer-reviewed publications (full list in section 14), of which those below have GPC co-authors.
Endpoint (OECD test guideline) | Approach | Publication | Status |
Bioconcentration in fish (TG 305) | Machine-learning q-RASAR | Pore, Pelloux et al. (2024), Journal of Hazardous Materials 479, 135725 | Published |
Fish early-life-stage toxicity (TG 210) | Intelligent consensus prediction (combined QSAR models) | Pore, Pelloux, Bergqvist et al. (2025), Aquatic Toxicology 279, 107216 | Published |
Reproductive and developmental toxicity (TG 421/422) | Read-across-driven classification | Chatterjee, Pore, Szepesi and Roy (2025), SAR and QSAR in Environmental Research 36(3), 247–270 | Published |
Repeated-dose toxicity (TG 407, 408, 422) | Machine-learning QSAR, with experimental validation | Pore, Szepesi and Roy (2026), Chemical Research in Toxicology 39(4), 544–565 | Published |
In vivo micronucleus (TG 474) | Integrated machine learning (structural, metabolic and experimental information) | Das, Szepesi and Roy, Toxicology in Vitro | Submitted October 2026 |
In vivo comet assay (TG 489) | In silico modelling | Das, Szepesi and Roy | In preparation |
NAMs in genetic toxicology | Review of regulatory screening and NAM application | Das, Szepesi, Ojha and Roy, book chapter, Comprehensive Computational Toxicology (Elsevier) | Due 30 November 2026 |
GPC authors in bold. The models cover endpoints for which animal studies are normally required under REACH, and they feed directly into the testing-avoidance methodology (below) and the chemical safety profiles (Objective 3).
Principle. As Lead Registrant for more than 200 substances, GPC could simply commission the studies ECHA requests. It chooses to research whether they can be avoided, accepting a smaller commercial gain from study management.
6. Objective 3 – Group-based chemical safety assessment (new stream)
Why it matters. Assessing chemicals one by one is slow and drives repeated testing. Regulators increasingly expect grouping and read-across, but there are few well-documented reference cases showing how group assessment can be done robustly.
Research questions. How can chemicals be grouped by functional group and chemistry category so that data and conclusions can be shared within a group? What evidence makes a group assessment scientifically defensible?
Activities
- Handbook of Chemical Safety Profiles: a reference set of detailed, endpoint-by-endpoint safety profiles for about 150 chemicals, by December 2028. More than 30 profiles completed in 2025–2026. Each profile records the reliability of the data, the conclusion per endpoint, the remaining data gaps and any scientific basis for avoiding studies. Lead: Dr Anders Bergqvist and Dr Zsuzsanna Szepesi.
- Group assessment approaches: functional-group and chemistry-category toxicological profiles developed from the Handbook, with the profiles serving as case examples for group chemical safety assessment.
Expected outcome. A documented basis for group assessments that reduces the need for future studies across whole groups of substances.
7. Objective 4 – Data exchange and data mining
Why it matters. Relevant data often exist but are scattered across databases, literature and private data owners, so studies are repeated unnecessarily.
Activities and results
Platform | Period | What it does | Results |
Datagap (data mining) | May 2026 – launch Dec 2026; internal use since Jul 2026 | Searches more than 150 sources (databases, literature, data owners, handbooks, government publications), presents data by required endpoint per region, and identifies data gaps and studies needed per submission | Platform in internal use |
Chemsortia (data exchange and sale) | Mid-2024 – launch May 2027 | Centralises information on existing study data and matches it to companies’ requirements; covers data GPC holds, data with other owners, and new study design where no data exist | Data on more than 10 substances shared, avoiding 25–80% of new studies for those substances |
The platforms use the regulatory science of Objective 1 (what is required) and the methods of Objectives 2 and 3 (when existing data are enough).
8. Objective 5 – Cost-efficient and realistic biodegradation testing
Why it matters. Biodegradation kinetics experiments are expensive and slow, and standard tests use single substances at high concentrations, unlike real effluents where many chemicals degrade together at low concentrations.
Aims.
- Minimise the cost of studies on degradation kinetics by testing many chemicals together in one experiment.
- Simulate real situations by testing complex effluents and discharge samples.
- Cut the cost and time of degradation kinetics experiments by at least 50–60% compared with testing substances individually.
Activities and results
Activity | Partner | Period | Key result |
Biodegradation kinetics of chemicals in mixtures (co-financed research) | DTU | 2019 – 2022 | GPC defined the research question (heterogeneous mixtures of industrial chemicals); Birch et al. (2023), Biodegradation and MethodsX; one paper discussed at ECHA and the EU PBT expert group |
Biodegradation of large numbers of discharged chemicals (PhD project) | DTU, IIT Bombay | 15 Feb 2025 – Q3 2028 | GPC proposed extending the method from GC-MS to LC-MS, so that higher-melting substances can be included; laboratory work at DTU (from Oct 2025) and in river water at IIT Bombay (from Feb 2026) |
[Add how the 50–60% saving will be measured, e.g. cost and duration per substance in mixture testing versus individual testing.]
9. Emerging area: sustainability and ESG
Intern projects in 2024–2026 analysed more than 1,000 sustainability (BRSR) reports and reviewed climate regulation and ESG reporting, enabling GPC to offer Sustainability Services. The next step is to turn the data analysis into a documented research output.
10. Collaboration model
Partner | Type | Period | Objective |
DTU | Co-financed research; co-funded PhD | 2019–2022; 2025–2028 | 5 |
IIT Bombay | PhD partner | 2025–2028 | 5 |
Jadavpur University | Joint modelling research | 2023–2026 | 2, 3 |
AFSA Collaboration | Industry–NGO research consortium | 2023–2027 | 2 |
Lund University / LTH | Joint research, theses, internships | 2019–2026 | 1, 4 |
KTH, RISE, industry partners | Vinnova project | 2024–2025 | 1 |
S&P Global | Joint PFAS white paper | 2026 | 1 |
Students linked to R&D (2025–2026): PFAS intern and thesis student (LTH); software and AI interns supporting the platforms.
11. How we work
- Identify a gap from practice: a test that gives no meaningful result, a missing model, a regulation that is hard to apply.
- Formulate the research question and how the answer will be used.
- Choose partners and agree roles, budget and publication rights in writing.
- Plan with milestones, and keep project files (agreements, minutes, drafts, data).
- Do the research, reviewing results continuously.
- Publish in peer-reviewed journals, reports and conferences.
- Transfer results into methods, platforms and services.
12. Timeline
Year | Milestones |
2019 | DTU mixture biodegradation project and DTU difficult-substances project start |
2022 | DTU mixture project concludes; regulatory science research and testing-avoidance methodology start |
2023 | Jadavpur collaboration on read-across and QSAR models starts; DTU mixture papers |
2024 | Intrims development and first user tests (40+ users); Chemsortia development starts; Macmillan et al. (AFSA); Vinnova biologics project |
2025 | DTU PhD starts (Feb); DIM project; Intrims patent filed (Jul); second Intrims user tests (25+ users); Global Inventory Report; Asia Chemicals Forum survey; Jadavpur papers |
2026 | DIM paper (Aug); repeated-dose toxicity paper; Datagap (May–Dec); PFAS thesis and white paper; biologics manuscript; patent published (Sep); Global Chemical Regulatory Framework (Nov) |
2027 | Intrims and Chemsortia launch; DIM next phase; PFAS method completed |
2028 | DTU PhD concludes; Handbook of about 150 chemical safety profiles completed |
13. Managing and documenting R&D
- Monthly time recording per person, by project and by R&D versus other work.
- One project register with research question, objective, partners, dates, staff and status.
- Project files with agreements, plans, minutes, correspondence, drafts, data and results.
- Development records for Intrims, Datagap and Chemsortia: versions, test rounds and change logs.
- Publication and acknowledgement policy agreed with partners at the start.
- Annual R&D report summarising objectives, projects, outputs and time spent.
French 
English 
Spanish 
German 
Korean 
Chinese 
Turkish 
Italian