Publications

Abstract : The SARS-CoV-2 virus has demonstrated its ability to adapt and spread in various environments, making it a challenging target for identification and prediction. While current studies in the field concentrates on utilization of transcriptome sequence classification to identify the virus, circular RNAs (circRNAs) have shown potential as a diagnostic marker for viral diseases. These single-stranded, covalently closed RNA molecules possess unique features such as RNA binding capacity and expression regulation, making it a promising source for potential biomarkers to create a new classification model. In this study, we propose a circRNA-based classification model utilizing the dna2vec algorithm to extract distributed representations of variable-length k-mers, combined with classical machine learning algorithms. The results demonstrate superior performance of the model, with Random Forest classifier achieving an accuracy of 99.99%, highlighting the efficacy of circRNA-based classification for SARS-CoV-2 identification and the potential of circRNAs as diagnostic markers for viral diseases.