News

Endocrine disruptors

LLMs & Akademie Fresenius Conference: Endocrine Disruptors

LLMs are being explored as supportive tools for predicting endocrine disruption (ED) potential of chemicals, complementing Quantitative Structure–Activity Relationship (QSAR) models in toxicological risk assessment.
Our colleague Arianna Bassan has been invited to present on this topic at the 16th International Akademie Fresenius Conference: Endocrine Disruptors16th International Akademie Fresenius Conference: Endocrine Disruptors on 25 and 26 November in Dusseldorf/Germany, where she will deliver a talk titled:

“From QSAR Predictions to LLM Evidence Summarization: Using AI for Endocrine Disruptor Assessment.”

In her presentation, Arianna will discuss how artificial intelligence can be integrated into the assessment of endocrine disruption potential, focusing on:
🔹 Predictive AI - Application of QSAR and machine learning models to predict endocrine activity in safety assessment
🔹 Generative AI - The role of LLMs in summarizing scientific evidence, literature, and mechanistic data to support regulatory decision-making
🔹 Explainability & Transparency - Addressing accountability, transparency, and interpretability as essential pillars for trustworthy AI in toxicology

At Onesum, our team is actively exploring how LLMs can support toxicological risk assessment by helping to identify relevant scientific literature, extract key data on hormonal pathways, and summarise complex evidence related to estrogen, androgen, thyroid, and steroidogenesis (EATS) modalities.

We recognise that these models have important limitations that must be carefully addressed. For this reason, our proprietary platform integrates LLMs within structured, human-in-the-loop workflows that separate data retrieval, extraction, and analysis, ensuring transparency, traceability, validation, and accountability.

LLMs are proving valuable for evidence mapping, literature synthesis, and supporting expert-driven weight-of-evidence conclusions. Ultimately, toxicologists retain full responsibility for final decisions, reinforcing that trust in AI-enabled ED prediction relies on accuracy, explainability, and rigorous expert review.

 

🌟Interested in collaborating?🌟

If you would like to discuss how LLMs can support your endocrine disruptor assessment projects, explore collaborative opportunities, or learn more about our approach, we would be happy to connect. Get in touch with the Onesum team to start the conversation.

www.onesumportal.com

 

Let’s work together

Whether you want to explore the Onesum Platform or engage our expert consultants, we’d love to learn about your thoughts and challenges.

Image

Onesum Portal

contact@onesumportal.com