ASNP: A Personalized Alternative Splicing Neoantigen Discovery Pipeline

Abstract

Neoantigen is important for immunotherapy. At present, neoantigen prediction is designed mainly based on gene mutations, but there are no methods to predict neoantigens from alternative splicing. We developed a novel ASNP (a personalized alternative splicing neoantigen discovery) pipeline based on alternative splicing information. ASNP uses a new selection strategy to capture potential neoantigen candidates from short peptides generated by alternative splicing, and perform HLA binding affinity prediction to identify high-affinity neoantigen candidates. Then, we applied the pipeline to non-small cell carcinoma and found 21302 potential HLA-peptide pairs. Finally, we evaluated our prediction results through the IEDB database and found 4 experimentally confirmed epitopes among potential candidates. The application of ASNP can help find potential neoantigen candidates and quickly develop personalized immunotherapy for cancer patients.