Return
Leveraging attention-based deep multiple instance and multiple task learning for improved neoepitope identification
DOI:10.1016/j.cels.2025.101404.png)
Abstract
En 中文
• New pan-allele MHC class I ligand and neoepitope predictor • Dynamically weights multi-allele interactions to eliminate scoring bias • Used ranking loss to enhance neoepitope prioritization • Outperforms existing methods and open-source code

