Assessing chemical toxicity across Eukaryota using multimodal transformers

bioRxiv (Cold Spring Harbor Laboratory)

Abstract

Abstract Biodiversity is globally threatened by chemical pollution, yet toxicity data remain unavailable for millions of species and tens of thousands of chemicals, severely limiting our ability to assess ecological impacts. Here we present TRIDENT-2, a multimodal artificial intelligence model for predicting chemical toxicity across evolutionarily diverse eukaryotic species. Trained on 560,780 toxicity assays spanning 82,775 chemicals, 6,793 species, and multiple exposure scenarios, TRIDENT-2 accurately predicts toxicity across Eukaryota with an average median absolute error ranging from 1.76 to 3.80. By jointly learning from chemical, biological, and experimental information, it remains accurate across broad chemical and taxonomic distances, allowing for toxicity assessment for species and chemicals beyond the current experimental evidence. Our findings demonstrate that artificial intelligence can help overcome longstanding data limitations in ecotoxicology, paving the way for improved decision-making and reducing chemical impacts on biodiversity and ecosystems.

Authors 7

1. Styrbjörn Käll

   Science for Life Laboratory · Chalmers University of Technology · University of Gothenburg

   Affiliation as printed

   Department of Mathematical Sciences, Chalmers University of Technology/University of Gothenburg, Gothenburg, Sweden

   Knightec Group, Gothenburg, Sweden

   Science for Life Laboratory (SciLifeLab), Gothenburg, Sweden

2. Mikael B. Gustavsson

   University of Gothenburg

   Affiliation as printed

   Department of Biological and Environmental Sciences, University of Gothenburg, Gothenburg, Sweden

3. Patrik Svedberg

   University of Gothenburg

   Affiliation as printed

   Department of Biological and Environmental Sciences, University of Gothenburg, Gothenburg, Sweden

4. Mercedes Dalman

   Chalmers University of Technology · University of Gothenburg

   Affiliation as printed

   Department of Mathematical Sciences, Chalmers University of Technology/University of Gothenburg, Gothenburg, Sweden

   Department of Medical Biochemistry and Cell Biology, University of Gothenburg, Gothenburg, Sweden

5. Jens Henriksson
   Affiliation as printed

   Knightec Group, Gothenburg, Sweden

6. Thomas Backhaus Aachen Chair for Ecotoxicology and Environmental Risk Assessment Institute for Environmental Research

   University of Gothenburg · RWTH Aachen University

   Affiliation as printed

   Chair for Ecotoxicology and Environmental Risk Assessment, Institute for Environmental Research, RWTH Aachen University, Aachen, Germany

   Department of Biological and Environmental Sciences, University of Gothenburg, Gothenburg, Sweden

7. Erik Kristiansson corresponding

   Science for Life Laboratory · Chalmers University of Technology · University of Gothenburg

   Affiliation as printed

   Department of Mathematical Sciences, Chalmers University of Technology/University of Gothenburg, Gothenburg, Sweden

   Science for Life Laboratory (SciLifeLab), Gothenburg, Sweden