O12 - Annotation of Canine Cancer Genomic Biomarkers Through Comparative Analysis of Human Mutations in COSMIC

Abstract: Genomic diagnostics are routinely used in human cancer medicine to guide diagnosis, treatment and prognostication. Growing evidence supports that human and canine cancers share genomic and clinical similarities. However, mutation-based biomarkers from canine data are less well-established than those utilized in human precision medicine.

Our objective was to bridge the canine precision medicine knowledge gap via “caninization” of the largest database of human cancer mutations, the Catalogue of Somatic Mutations in Cancer (COSMIC), and consume these data into a canine cancer knowledgebase, Vidium Insight.

To identify canine equivalents of human COSMIC mutations, cross-species conversion was performed via both genomic liftover and protein alignment, considering evolution conservation score and sequence homology. Known/predicted pathogenic mutations in COSMIC Cancer Gene Census (CGC) genes were selected for “caninization” as a proof-of-concept.


Of 1763970 total mutation records in 717 CGC genes in COSMIC (v95), 295439 unique, pathogenic mutations in 707 CGC genes across ~200 tumor types were initially converted. Histologically similar canine cancers exist for most primary human cancer types, thus these inferences can be informative for dogs. Of 707 human CGC genes, 95% have canine orthologs and are highly conserved among species. Canine equivalents were identified for ~80% of human mutations.


Most of the nearly 3 million documented pathogenic human cancer mutations have canine counterparts. Thus, systematic “caninization” of these human cancer mutations with established biomarker associations will not only expand genomic knowledge of canine cancers in research settings, but will also improve utilities of genomic diagnostic tools in canine cancer patient care.