Abstract LB344: CertisOI Assistant™ - Accelerating Preclinical Cancer Model Selection with Generative AI

Abstract Pre-clinical oncology researchers face significant challenges when identifying suitable tumor models. Suboptimal model selection can impede the translational success of preclinical findings to clinical applications. Traditional online databases rely on rigid filtering systems that may not accommodate the complex and specific criteria researchers require. To address this critical need, Certis Oncology developed CertisOI, a Generative AI-powered platform that transforms how researchers discover and analyze cancer tumor models for their studies. CertisOI functions as an interactive research assistant, enabling users to articulate complex tumor selection criteria in natural language, generate dynamic queries, and retrieve actionable insights from a comprehensive database of pre-clinical models. The system integrates Certis Oncology's proprietary patient-derived xenograft (PDX) models with the Cancer Cell Line Encyclopedia (CCLE) database, providing information on tumor characteristics, patient demographics, growth dynamics, pharmacological studies, genetic analyses, and drug efficacy predictions. It also integrates with the company’s predictive AI platform, which uses a machine learning algorithm trained on experimental data to predict drug response based on chemical and biological features. This adds a level of sophistication to preclinical model selection, as it allows researchers to select models most likely to be responsive or resistant to FDA-approved therapies. What sets CertisOI apart is its ability to conduct sophisticated real-time analysis through an integrated coding environment, presenting insights through multiple modalities including visualizations, interactive charts, tables, and downloadable datasets. This dynamic approach allows researchers to iteratively refine their model selection criteria based on emerging insights, leading to more precise model identification and ultimately accelerating the path to clinical trials. By bridging the gap between complex research requirements and vast pre-clinical data, this platform represents a significant advancement in oncology research tools, enabling more efficient and effective pre-clinical model selection. Citation Format: Luke Jervis, Yuan-Hung Chien, Warren Andrews, Raffaella Pippa, Long Do. CertisOI Assistant™ - Accelerating Preclinical Cancer Model Selection with Generative AI [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 LB344.