Somatic mutations can affect key domains of cancer genes. These mutations, associated with cancer progression and resistance to therapy, exist in restricted regions of the genome, termed mutational hotspots. Additionally, actionable mutations, in which an approved or investigational agent is available to target a pathway activated by the mutation, exist in an even more restricted set of these genomic regions. While most available clinical assays interrogate one or only a few commonly mutated loci in cancers, two published clinical assays, SNaPShot [3] and OncoMap [4], respectively target 38 mutations in 8 genes by single base extension assays and approximately 400 mutations in 33 genes by mass spectrometry. Although these assays have been extensively tested on clinical samples and are available to clinicians, they have not been thoroughly evaluated on heterogeneous tumor samples. Technological advances in DNA sequencing clearly offer an important solution to the problem of analyzing heterogeneous samples. Massively parallel sequencing enables the analysis of independent, clonal, DNA molecules [5,6] and has been used early on to digitally measure the presence of low prevalence mutations in complex DNA mixtures