Identifying and characterizing somatic variants in deep genome sequence data from tumor samples remains challenging and time-consuming. Of special interest in cancer research and diagnostics is the detection and annotation of rare subclonal somatic variants found only in a small proportion of primary tumor cells. Such variants can drive tumor spread and recurrence, but are often neglected in choosing treatments.

Currently, few tools reliably distinguish such rare subclonal variants from sequencing errors. And even among real somatic variants, drivers (of tumor growth, spread, or resistance) are hard to distinguish from passengers. Doing so entails integrating diverse information on variants, genes, pathways, cancer-relevant phenotypes, and treatments (including insights on population allele frequencies and broader evolutionary conservation; known/likely effects on gene product structure, function, expression, and interaction; and relations among gene products, phenotypes, and drugs). Software for effectively integrating such data in light of genomic variation in samples, to highlight relevant findings through clear visualization, has been a pressing need.

Here we present an end-to-end analysis workflow for finding and functionally characterizing rare subclonal variants, using the newly developed CLC Cancer Research Workbench to feed the interpretive platform of Ingenuity Variant Analysis, to identify cancer driver mutations in paired tumor/normal samples. We will show new interesting results from this analysis, which were not shown beforehand on this publicly available cancer dataset (Case Reports in Oncological Medicine, Volume 2013 (2013), Article ID 270362) from a patient with massive acinic cell carcinoma.