ESTIV New Award, EMA qualification for Certara, BioBricks : New open-source Data-Hub, Novel Antibiotics by GenAI, OwkinZero : Biological reasoning blind spot in current LLMs
News on non-animal methods
SEPT. 01 - 05, 2025NEWS, REPORTS & POSITION STATEMENTS
1. Bringing scientific methods into regulatory guidelines
Many innovative, animal-free methods never make it into official regulatory frameworks — not because they lack scientific value, but because navigating the path to acceptance is complex. A new article highlights the challenges and opportunities of getting methods adopted into international guidelines, such as those from the Organisation for Economic Co-operation and Development (OECD) or International Organization for Standardization (ISO).
The article outlines the structured process of validation, stakeholder engagement, and approval that is essential for broader use in safety testing and policy. The authors also introduce tools from the NanoHarmony project, including a free Test Guideline Process Mentor that helps researchers understand, and plan, each step of the process.
Read more in Frontiers Research Metrics and Analytics
2. South Korea begins construction of national animal-free testing facility — completion set for 2026
The Ministry of Environment in South Korea has started construction of the nation’s first dedicated animal-free testing facility — part of ‘a national goal of replacing over 60% of toxicity testing with alternative methods by 2030’.
Vice-Minister of Environment Lee Byung-hwa said “The groundbreaking of this animal alternative testing facility is an important step toward opening an era in which animal welfare and public health are protected together”, and added, “we will do our best to promote related policies so that our country can leap forward as a world-class leader in testing technology”. With construction launched in May 2025, the facility is due to open in October 2026.
Read more details on the animal-free testing facility
INTERVIEWS, NOMINATIONS & AWARDS
3. Summer replay of Season 3 of Café Scientifique, webinar series where science, policy, and imagination meet
Café Scientifique is a series of regular webinars delivered by PrecisionTox project partners and collaborators on topics related to the project subject domain.
Season 3 delves into key regulatory and scientific breakthroughs, such as : Key Areas of Regulatory Challenge (KARC) with Tomasz Sobański, European Chemicals Agency ; TOP of the TOX with Laura Holden, Birmingham Law School ; OoC : Advancing the safety assessment of chemicals in human health and environment with Janire Saez Castaño, Ikerbasque & UPV/EHU… These sessions examine where science meets regulation — and how Next-Generation Risk Assessment (NGRA) and New Approach Methodologies (NAMs) are reshaping the future of chemical safety.
4. Nominations for Lush Prize 2026 are now open
The Lush Prize is the largest prize fund in the non-animal testing sector, with a prize fund of £250,000 to support initiatives to end or replace animal testing. The Lush Prize has five main prize categories : Science, Training, Public Awareness, Political Advocacy, and Young Researcher. In addition there are Recognition Awards, for Political Achievement, Major Science Collaboration and the Andrew Tyler Award for outstanding contribution towards ending animal testing.
5. ESTIV launches the Liebsch – Spielmann Award for contributions to the validation of alternative methods
Named in honour of Dr. Manfred Liebsch and Prof. Dr. Horst Spielmann, two pioneers in the field of alternative methods and their validation, this new award will recognise outstanding contributions to the scientific and regulatory advancement of non-animal testing approaches.
The award was officially introduced to the international scientific community during the 13th World Congress on Alternatives and Animal Use in the Life Sciences in Rio de Janeiro, Brazil (August 2025). The first recipient of the ESTIV Liebsch – Spielmann Award will be formally honoured during the ESTIV2026 Congress in Maastricht, the Netherlands. With this initiative, ESTIV (European Society of Toxicology In Vitro) continues its long-standing commitment to supporting the development, validation, and regulatory acceptance of innovative, human-relevant, and ethical methods in toxicology.
Read the announcement on LinkedIn
TOOLS, PLATFORMS, CALLS
6. WHO : Guidelines on the replacement or removal of animal tests for the quality control of biological products
Growing recognition that animal tests are generally inferior to appropriately developed in vitro assays for the routine quality control of biological products is becoming a key driver of a shift towards quality control schemes that avoid the use of animals altogether. The WHO published a draft document for the purpose of inviting comments and suggestions on proposed Guidelines on the replacement or removal of animal tests for the quality control of biological products to a broad audience and to improve transparency of the consultation process, which will then be considered by the WHO Expert Committee on Biological Standardization (ECBS).
Written comments proposing modifications to this text MUST be received in English by 19 September 2025 in the Comment Form available separately. The outcome of the deliberations of the ECBS will be published in the WHO Technical Report Series. The final agreed formulation of the document will be edited to be in conformity with the second edition of the WHO style guide (KMS/WHP/13.1).
7. BioBricks : New open-source Data-Hub to identify toxic chemicals before they reach market
Researchers at the Johns Hopkins Bloomberg School of Public Health, in collaboration with Insilica LLC and a team of U.S. and international collaborators, are launching BioBricks.ai — an open-source data platform designed to supercharge our ability to predict the toxic effects of chemicals by powering AI models with highly comprehensive, standardized, and easily computed data sets.
This new resource gives researchers free access to a curated library of over 90 biological and chemical datasets. The dataset covers millions of chemicals, from environmental toxins to drug data – all publicly available but never before compiled and standardized – that are bundled into easy-to-use “BioBricks” downloadable from biobricks.ai and GitHub. The dataset is described in a peer-reviewed paper published online August 13 in the journal Frontiers in Artificial Intelligence.
Read the publication in Frontiers in Artificial Intelligence
INDUSTRY, BIOTECH & PARTNERSHIPS
8. Certara Simcyp® Simulator receives EMA qualification opinion for PBPK modeling
Certara, Inc., a global leader in biosimulation, announced that the European Medicines Agency (EMA) has formally qualified the Simcyp® Simulator for use in regulatory submissions across the EU to assess drug-drug interaction (DDI) risk.
The qualification opinion covers various scenarios (3 COUs) where Simcyp simulations can replace clinical DDI studies using a range of verified substrates and perpetrators integrated within the platform.
The COUs span 6 CYP enzymes and 2 inhibition mechanisms, thereby significantly reducing the number of clinical studies a sponsor has to conduct. “This first of its kind qualification by EMA regulators of a new approach methodology further demonstrates the scientific value biosimulation provides to drug development scientists,” said William F Feehery, Chief Executive Officer.
9. Silicon Valley’s Molecular Devices Brings AI to Brain Organoid Culture
Molecular Deviceshas released an upgrade to its CellXpress.ai Automated Cell Culture System, adding rocking incubation and AI-driven workflow automation for brain organoid culture. The feature is designed to reduce the high manual workload of organoid maintenance and to support long-term, reproducible experiments for neurodegenerative disease research.
With the new upgrade, rocking incubation maintains nutrient flow within the wells, while AI-guided scheduling automates feeding, media changes, and imaging. The system operates continuously on a 24/7 schedule and records all events, including digital microscopy data, in a unified software environment. This provides reproducibility across assays, supports multiple stem cell lines, spheroids, or organoids in parallel, and maintains standardized protocols that reduce human error.
10. FDA grants de novo authorization to ArteraAI’s multimodal AI prostate cancer risk software
The U.S. Food and Drug Administration has grantedDe Novo authorization to ArteraAI Prostate, a multimodal artificial intelligence (MMAI) software developed by Artera, Los Altos, California headquartered precision medicine company, designed to assist clinicians with risk-based decisions in localized prostate cancer. The decision follows the product’s earlier Breakthrough Device Designation and allows its use as a regulated medical device in qualified pathology laboratories.
Artera’s multimodal artificial intelligence (MMAI) platform combines digitized pathology slides with clinical data to deliver predictive and prognostic insights for personalized cancer care. Built with diverse patient cohorts, including representative African American populations, it preserves tissue by using existing biopsy slides and provides results within days. The system is designed to augment, not replace, physicians, reducing uncertainty in treatment benefit and enabling AI biomarker development across cancer types.
SCIENTIFIC DISCOVERIES & PROTOCOLS
11. Cell Reprogramming : OpenAI and Retro Biosciences Report 50x Pluripotency Marker Expression Gains
OpenAI and Retro Biosciences have released new findings from a joint effort in AI-driven protein engineering. Using GPT-4b micro, a biology-specialized variant of GPT-4o, the team designed two modified versions of Yamanaka factors, RetroSOX and RetroKLF, that showed over 50-fold increases in pluripotency marker expression and improved DNA repair activity during cell reprogramming.
Over 30% of AI-generated RetroSOX variants showed higher reprogramming activity than native SOX2, with some differing by more than 100 amino acids, while nearly 50% of RetroKLF variants exceeded manually engineered versions. The approach was also tested in mesenchymal stromal cells from older adult donors, where over 85% of cells activated endogenous pluripotency genes within 12 days. Cells reprogrammed with the AI-designed proteins demonstrated full differentiation potential and maintained normal chromosomal structure.
12. Generative AI designs novel antibiotics against MRSA and drug-resistant Gonorrhea
Researchers at MIT and the Broad Institute have reported generative artificial intelligence systems capable of designing entirely new antibiotics, including a compound active against methicillin-resistant Staphylococcus aureus (MRSA). The study, published August 14 in Cell, describes two lead molecules—NG1 and DN1—produced by deep learning models trained on a dataset of ~40,000 small molecules with known antibacterial profiles.
Both models employed predictive filters developed at MIT to estimate whether proposed molecules could be synthesized. The code for the generative models has been released openly. James Collins, Ph.D., Broad Institute, who co-led the project, has previously worked on AI-guided antibiotic discovery, including the 2020 identification of halicin.
13. OwkinZero : Addressing the biological reasoning blind spot in current LLMs
While large language models (LLMs) are rapidly advancing scientific research, they continue to struggle with core biological reasoning tasks essential for translational and biomedical discovery. To address this limitation, Owkin created and curated eight comprehensive benchmark datasets comprising over 300,000 verifiable question-and-answer pairs, each targeting critical challenges in drug discovery including target druggability, modality suitability, and drug perturbation effects.
The result uncovered evidence of a key aspect of generalization : specialist models trained on a single task consistently outperform their base models on previously unseen tasks. This study represents a significant step toward addressing the biological reasoning blind spot in current LLMs, demonstrating that targeted reinforcement learning on carefully curated data can unlock generalizable performance in specialized models, thereby accelerating AI-driven biological discovery.
14. Coordinated development of immune cell populations in vascularized skin organoids from human induced pluripotent stem cells
The research team at University of Queensland’s Frazer Institute used stem cells to create a replica of human skin, complete with blood vessels, capillaries, hair follicles, layers of tissue and immune cells. Currently, there is an unmet challenge to generate vascularized, functional skin models. In this study, human fetal placental endothelial colony forming cells (ECFC) are incorporated into human induced pluripotent stem cell (hiPSC)-derived skin organoids (SKO), forming capillary-like structures and generating endothelialized SKOs.
Dr Abbas Shafiee said the skin model, that was 6 years in the making, would be transformative for skin graft transplants, wound healing and studying skin disorders. “This is the most life-like skin model that’s been developed anywhere in the world and will allow us to study diseases and test treatments more accurately,” Dr Shafiee said. The research received funding from Metro North Health and was carried out at the Translational Research Institute.
Read the publication in Wiley Advanced Healthcare Materials
