FusionPro, a Versatile Proteogenomic Tool for Identification of Novel Fusion Transcripts and Their Potential Translation Products in Cancer Cells

Fusion proteoforms represent all protein products that can be generated by translation of fusion transcripts. FusionPro is developed as a sensitive tool for detecting and annotating fusion transcripts and proteoform by analyzing RNA-Seq data. In this study, we found the evidence of fusion proteoforms present in MS/MS data and analyzed their translational patterns based on the sequence information which is obtained from FusionPro. Our pipeline will facilitate the proteogenomic identification of fusion-derived peptides and characterization of various oncogenic fusion proteoforms. Fusion proteoforms are translation products derived from gene fusion. Although very rare, the fusion proteoforms play important roles in biomedical science. For example, fusion proteoforms influence the development of tumors by serving as cancer markers or cell cycle regulators. Although numerous studies have reported bioinformatics tools that can predict fusion transcripts, few proteogenomic tools are available that can predict and identify proteoforms. In this study, we develop a versatile proteogenomic tool FusionPro, which facilitates the identification of fusion transcripts and their potential translatable peptides. FusionPro provides an independent gene fusion prediction module and can build sequence databases for annotated fusion proteoforms. FusionPro shows greater sensitivity than the available fusion finders when analyzing simulated or real RNA sequencing data sets. We use FusionPro to identify 18 fusion junction peptides and three potential fusion-derived peptides by MS/MS-based analysis of leukemia cell lines (Jurkat and K562) and ovarian cancer tissues from the Clinical Proteomic Tumor Analysis Consortium. Among the identified fusion proteins, we molecularly validate two fusion junction isoforms and a translation product of FAM133B:CDK6. Moreover, sequence analysis suggests that the fusion protein participates in the cell cycle progression. In addition, our prediction results indicate that fusion transcripts often have multiple fusion junctions and that these fusion junctions tend to be distributed in a nonrandom pattern at both the chromosome and gene levels. Thus, FusionPro allows users to detect various types of fusion translation products using a transcriptome-informed approach and to gain a comprehensive understanding of the formation and biological roles of fusion proteoforms.