The vast majority of genes is alternatively spliced and many such spliced isoforms are specifically associated with cancer progression and metastasis. The splicing pattern of numerous genes is altered as cells move through the oncogenic process towards a more aggressive invasive cancer phenotype (Kozlovski et al. 2017; Pedersen and Jensen 2017; Shilo, Siegfried, and Karni 2015). Moreover, splicing factors might be de-regulated in cancer, and in some cases aberrant isoform might appear and drive aspects of the disease progression. Targeting those specific cancer isoforms might provide novel and less toxic therapeutic molecules. However, it is very challenging to experimentally detect and validate those isoforms on the protein level (Barrett et al. 2015). Here, we present a novel pipeline for the identification of rare isoforms likelihood to yield a folded protein from RNAseq data. These results would identify new genetic biomarkers associated to cancer risk, prognosis and response to standard treatment.