Abstract 3690: Biomarker Navigator - empowering the augmented scientist in biomarker discovery

Abstract The advent of large language models (LLMs) is poised to revolutionize every domain of the pharmaceutical industry, from early-stage drug development to clinical trial design. By harnessing both public datasets and scientific literature, LLMs offer unparalleled access to critical knowledge for researchers, while proprietary datasets provide a unique competitive edge, enabling organizations to derive exclusive insights and drive innovation. To excel in biomarker development within large organizations, two critical goals must be prioritized: maximizing the extraction of insights from proprietary datasets and ensuring the seamless dissemination of knowledge across the organization. Large Language Models (LLMs) facilitate these goals by connecting expertise across diverse scientific domains and delivering knowledge in natural language. This enables the creation of an “augmented scientist” empowered to make data-driven decisions and generate hypotheses through the integration of cross-disciplinary knowledge. The Biomarker Navigator program exemplifies AstraZeneca’s commitment to this augmented scientist vision through three core components: 1) an automated analysis framework for biomarker identification. 2) a comprehensive repository of standardized, detailed biomarkers insights. 3) an intuitive interface for fast, seamless knowledge access. AstraZeneca is expanding the Biomarker Navigator by integrating public datasets to make it accessible to the broader scientific community, with an initial focus on renal cell carcinoma (RCC) and melanoma. This dataset expansion includes whole exome sequencing (WES), copy number alterations (CNA), RNA sequencing (RNA-seq), and clinical metadata. The resulting system empowers researchers to seamlessly transition from a high-level overview of the biomarker landscape—within or across datasets—to detailed insights about a single biomarker, ensuring no information is lost. By combining an intuitive user interface with LLM-driven insights, this platform showcases the transformative potential of AI in biomarker discovery, paving the way for a robust, open-source framework for the scientific community in the near future. Citation Format: Ioannis Kagiampakis, Mishka Gidwani, Iaroslav Korkodinov, Georgiy Ponomarev, Sun Gerald, Etai Jacob, Damian Bikiel. Biomarker Navigator - empowering the augmented scientist in biomarker discovery [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 3690.