Background: Pancreatic adenocarcinoma (PAAD) is referred to as an immunologically “cold” tumor that responds poorly to immunotherapy. A fundamental theory that explains the low immunogenicity of PAAD is the dramatically low tumor mutation burden (TMB) of PAAD tumors, which fails to induce sufficient immune response. Alternative splicing of pre-mRNA, which could alter the proteomic diversity of many cancers, has been reported to be involved in neoantigen production. Therefore, we aim to identify novel PAAD antigens and immune subtypes through systematic bioinformatics research. Methods: Data for splicing analysis were downloaded from The Cancer Genome Atlas (TCGA) SpliceSeq database. Among the available algorithms, we chose CIBERSORT to evaluate the immune cell distribution among PAADs. The TCGA-PAAD expression matrix was used to construct a co-expression network. Single-cell analysis was performed based on the Seurat workflow. Results: Integrated analysis of aberrantly upregulated genes, alternatively spliced genes, genes associated with nonsense-mediated RNA decay (NMD) factors, antigen presentation and overall survival (OS) in TCGA-PAAD revealed that PLEC is a promising neoantigen for PAAD-targeted therapy. We identified a C2 TCGA-PAAD subtype that had better prognosis and more CD8+ T-cell infiltration. We propose a novel immune subtyping system for PAAD to indicate patient prognosis and opportunities for immunotherapy, such as immune checkpoint (ICP) inhibitors. Conclusions: In conclusion, the present study used a transcriptome-guided approach to screen neoantigen candidates based on alternative splicing, NMD factors, and antigen-presenting signatures for PAAD. A prognosis model with guidance of immunotherapy will aid in patient selection for appropriate treatment.