Abstract LB108: CarD-T: Interpreting Carcinomic Lexicon via Transformers

Abstract The identification and classification of carcinogens is critical in cancer epidemiology. We introduce the Carcinogen Detection via Transformers (CarD-T) framework, combining transformer-based machine learning with probabilistic analysis to efficiently nominate potential carcinogens from scientific texts. Trained on 60% of established carcinogens, CarD-T correctly identifies all remaining known carcinogens and nominates ∼1,600 potential new carcinogens. Comparative assessment against GPT-4 reveals CarD-T's comparable precision (0.894 vs 0.903), and superior recall (0.857 vs 0.705), implying improved ability to classify carcinogens not in major databases. Additionally, CarD-T highlights 554 entities with disputing evidence, analyzed using Bayesian Probabilistic Carcinogenic Denomination (PCarD). The framework reveals significant shifts in research focus from chemical carcinogens to broader categories including environmental factors (18%), biological agents (10%), and emerging threats like COVID-19, supported by 577 publications since 2020. This framework enhances the agility of public health responses to carcinogen identification, setting a new benchmark for automated, scalable toxicological investigations. Citation Format: James (Jamey) ONeill, Parag A. Katira. CarD-T: Interpreting Carcinomic Lexicon via Transformers [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB108.