eccDNA2Ca: an ensemble deep learning framework for interpretable prediction of cancer-associated extrachromosomal circular DNA

eccDNA2Ca represents the first interpretable and publicly available computational framework specifically designed to prioritize cancer-relevant eccDNAs after experimental detection based on sequence features. By enabling scalable and biologically informed prioritization of eccDNAs, eccDNA2Ca provides a valuable resource for eccDNA functional studies, cancer biomarker discovery, and translational cancer genomics. Source code and datasets are freely available at: https://github.com/bread1006/eccDNA2Ca. An interactive web server implementing eccDNA2Ca is accessible at: http://43.138.143.50:5000/en.