Fully Convolutional Neural Network for Predicting Cancer-Specific CircRNA-MiRNA Interaction Sites

In recent years, the research of circRNAs in biomedical fields has gained momentum, particularly in cancer biology where they play a crucial role. CircRNAs regulate gene expression by adsorbing miRNAs and acting as ‘sponges’. Dysregulation of miRNAs has been observed in various cancer tissues, and co-expression of circRNAs with miRNAs has been noted in many cancer tissues. The co-expression of miRNAs with circRNAs may play an important role in regulating tumour growth. In this study, we predicted cancer-specific circRNA-miRNA interaction sites using a fully convolutional neural network model, and achieved high prediction accuracy by optimizing the model parameters. This study is significant for understanding the mechanism of circRNA’s role in cancer development and developing related therapeutic and diagnostic methods.