Background and Aim: There are over 400 genetically distinct breeds of dogs, each corresponding to a genetic isolate. The consequence of breeding practices is that most breeds naturally develop specific cancer types, reflecting the presence of predisposing alleles. This is interesting for lymphomas, as most human lymphoma subtypes are encountered in dogs, with some subtypes over-represented in specific breeds (Pastor et al., 2009; Rowell et al., 2011; Marconato et al., 2013). We focused on lymphoma occurring in large families of Bernese Mountain Dogs (BMDs) with the objectives to identify predisposing genetic regions and somatic alterations involved in lymphomagenesis. Materials and Methods: We collected blood and tumour samples, clinical and genealogical information of affected BMDs, using the French CaniDNA biobank at CNRS, Rennes. Histopathological diagnosis was performed from formalin-fixed paraffin-embedded tumour samples by veterinary pathologists F. N. and J. A. DNA and RNA were extracted using Macherey-Nagel® kits. Dog DNAs were genotyped on the canine Illumina®, and 170 000 single-nucleotide polymorphism (SNP) arrays and RNA of matched tumour and healthy tissues were sequenced (RNAseq) using Illumina technology, by the GIGA genomic facility, University of Liege. Data were analysed by the software crac/ chimCT (Philippe et al., 2013) to identify translocations. The fusion point was validated on tumour DNA and cDNA, by Sanger sequencing. Results: First, we demonstrated a familial segregation of lymphoma in a pedigree of BMDs. We performed a genome-wide association study using 63 lymphoma cases and 164 controls with the high-density SNP chips. Preliminary data showed a significant locus, associated with predisposition to lymphoma, on canine chromosome CFA 23. This locus does not correspond to any known locus for human lymphomas and thus represents an interesting candidate region. Second, while many chromosomal translocations are known to drive lymphomagenesis in humans, only little genetic data are available in dogs. We thus investigated somatic alterations through RNAseq on five lymphoma-affected dogs and identified a relevant translocation involving an immunoglobulin gene and a gene involved in cell cycle. Interestingly, this gene is known to be involved in translocations in human lymphomas, showing that similar genes are involved in both human and dog lymphomas. Conclusions: Based on the collection of different types of canine lymphomas in several predisposed breeds, we first identified a potential novel locus predisposing to lymphomas. We are exploring and reducing the region of interest by using other canine breeds. Second, we discovered a novel chromosomal translocation in dogs, homologous to a translocation in the same human lymphoma type. These data clearly showed that the same key pathways are involved in lymphomas in dogs and humans, justifying the setting up of clinical trials in dogs for the benefit of both humans and dogs.