Overview This project explores the integration of RNA 3D structural information with gene expression data to enhance cancer detection methodologies. By leveraging advanced machine learning techniques, particularly graph-based neural networks, the study aims to improve the accuracy of cancer classification by considering both the quantitative expression levels of genes and the spatial conformation of RNA molecules.
Background RNA molecules play a pivotal role in various biological processes, and their three-dimensional structures are crucial for their function. Aberrations in RNA folding can lead to dysfunctional gene expression, often associated with diseases like cancer. Traditional gene expression analyses focus on the quantity of RNA transcripts, potentially overlooking structural anomalies that could be indicative of pathological states. Incorporating RNA 3D structural data provides a more holistic view, potentially unveiling subtle biomarkers for early cancer detection.