The GF-NEO discovery platform unveils a [KQE][DG] sequence motif within fusion neoantigens in pediatric cancer

Gene fusions (GFs) are critical events in pediatric oncology, often serving as oncogenic drivers. However, fusion proteins and their derived neoantigens (GF-NEOs) remain underexplored for targeted immunotherapy. We developed ProteoFusioNEO, a computational tool for the <i>in silico</i> translation of transcriptomic data, analyzing 5,190 pediatric patients with cancer and 935 cell lines, yielding 382 and 446 fusion proteins. We highlight that GFs generate multiple translational outcomes, with 97% being in-frame in patients. Fusion junctions exhibit the sequence motif [KQE][DG], which partly reflects the nature of exon-exon junctions, albeit with additional hydrophilicity. Moreover, GF-NEOs' abundance may be shaped by the motif, offering insights into fusion protein biology. Finally, a multipronged validation strategy using <i>in vitro</i> and <i>in vivo</i> systems confirms the GF-NEOs presentation through mass spectrometry-based proteomics and immunopeptidomics. Multiple GF-NEOs encoded by two versions of the ETV6-RUNX1 fusion were validated, paving the way for targeted immunotherapy development.