A Heterogeneous Network Model Based Analysis and Detection of RNA Enabling Convolutional Neural Network

RNA is an important target in the research of genetic diseases and disorders since it is essential for the process of gene expression and control. Convolutional neural networks (CNNs) and other machine learning-based methods for analyzing RNA sequences have recently attracted more attention. But reliable analysis and detection are significantly hampered by the enormous variability of RNA sequences. In this research, a brand-new heterogeneous network model for CNN-based RNA analysis and detection is provided. This model accounts for the wide range of RNA sequence properties, such as length fluctuation, secondary structure, and chemical changes. A unified model that incorporates these features is created by combining 1D and 2D convolutional layers. Multiple datasets, including synthetic and real-world datasets, were used to assess proposed model's performance. The outcomes show that the model performs better in terms of accuracy, sensitivity, and specificity than current state-of-the-art approaches. Furthermore, the model's performance is greatly enhanced with the addition of 2D convolutional layers and attention processes, according to rigorous studies to assess the contributions of each component of the model. Overall, the heterogeneous network model offers a reliable and precise method for analyzing and identifying RNA sequences, which has significant ramifications for comprehending hereditary illnesses and creating targeted therapies.