TRA-Funding for Innovative Research: "Automatic construction of gene regulatory networks from scientific literature with LLMs"
The project aims to harness the capabilities of LLMs — such as those powering ChatGPT — to analyze vast amounts of scientific literature and extract meaningful data about gene interactions and regulations. LLMs and their application are one of the major research topics of Professor Lucie Flek. Now it has been awarded seed funding for one year by the TRA Modelling, full name being: "Mathematics, Modelling, and Simulation of Complex Systems". TRAs provide spaces for innovation in research and teaching that are tackling the interdisciplinary scientific, technological and societal challenges of the future. The TRAs are open to all researchers active at the University of Bonn and its partner institutions who are able to contribute in an interdisciplinary collaboration to the specific topics being studied.
The Role of Gene Regulatory Networks
Gene regulatory networks are crucial for understanding how genes interact within living organisms. Disruptions in these networks can lead to various health issues, with cancer being one of the most significant outcomes. As research into cancer genetics expands, the volume of published studies has surged, making it increasingly challenging for researchers to synthesize findings across different studies. Many existing papers focus on isolated genes or specific experimental conditions rather than comprehensive networks. Professor Flek's project aims to address this gap by using LLMs to extract and integrate partial networks described in literature, thereby creating a more holistic view of gene regulation in cancer.
Collaborative Efforts
The project will be conducted in consultation with Professor Holger Fröhlich, Head of the Biomedical Data Science Group at b-it. and Dr. Christiane Hellweg from the German Aerospace Centre (DLR). Professor Fröhlich specializes in statistical data mining and machine learning applications in biomedicine, while Dr. Hellweg’s expertise lies in radiation effects on organisms and their implications for cancer therapy. This collaborative approach highlights the importance of interdisciplinary research in tackling complex biomedical challenges.
The Future of AI in Biomedical Research
Professor Flek's work exemplifies the potential of Artificial Intelligence in advancing biomedical research: As AI continues to evolve, its application in fields like genomics and cancer research could revolutionize how scientists approach disease prevention and treatment. In a landscape where data is abundant but often fragmented, leveraging LLMs presents a promising solution for synthesizing knowledge across diverse studies. By automating the construction of gene regulatory networks from existing literature, this project could pave the way for new insights into cancer biology and ultimately contribute to more effective therapeutic strategies.