Leveraging LLMs for Automated Analysis of Biomedical Data

Large Language Models (LLMs), specifically trained to understand and process natural language queries, represent a transformative approach in biomedical data analysis, particularly for querying and interpreting complex datasets such as those available on cBioPortal. In this paper we implemented an automated workflow to analyze biomedical datasets and return relevant results, where LLMs are utilized for query analysis and code generation tasks. The significance of this work lies in its potential to facilitate access to complex biomedical data, reducing the barrier for researchers with limited computational skills, and to streamline the process of data analysis and interpretation. The use of LLMs ensures that queries can be formulated in intuitive natural language, making this workflow accessible to a broader audience. By automating the data analysis process, our workflow can be applied to a batch of data analysis tasks across a large number of datasets simultaneously, facilitating faster hypothesis generation and validation in biomedical research for accelerating discoveries in cancer genomics and personalized medicine.