Recent reports on subclonal events in cancer used non-standard experiments; they have either inferred subclonal status by looking for shared clonal events among several metastases from the same patient14, resorted to ultra-deep sequencing11 or sequenced very small numbers of single cells15–17. In contrast, tens of thousands of tumors are being sequenced at standard depths of 100–150x for exomes and 30–60x for whole genome as part of large scale cancer genome projects, such as The Cancer Genome Atlas (TCGA)1,2,7 and the International Cancer Genome Consortium (ICGC)18. In order to detect clonal and sub-clonal mutations present in these samples there is a need for a highly sensitive and specific mutation calling method. Although specificity can be controlled through subsequent experimental validation, this is an expensive and time-consuming step that is impractical for general application.
