How regulatory information is encoded in RNA structure is a longstanding problem in gene expression control. Here, I will describe our work bringing together linguistic concepts and information-theoretic frameworks to capture RNA structural elements that govern RNA dynamics. I will describe our application of this approach to aberrant splicing patterns observed in metastatic breast cancer. In this study, we discovered and deeply characterized a previously unknown RNA structural splicing enhancer (SSE) that drives oncogenic splicing. We have identified the associated RNA-binding protein that serves as the trans factor that interacts with this novel SSE. We have shown that this pathway drives metastatic progression in xenograft mouse models and that its activity is strongly associated with poor survival in patients. Finally, I will describe how we are leveraging the recent advances in machine learning to further expand our ability to identify structured regulatory elements in RNA.