We've done targeted NGS of EGFR pathway genes in tumors from colorectal cancer patients (retrospective study). To classify the mutant variants we identified as somatic, we used peripheral blood DNA as controls (the only control available to us), basically following the usual guidelines that if you find it in the control, it is probably just a germline polymorphism. Surprisingly though, even established and highly characterized canonical mutants such as KRAS G12D or KRAS Q61R showed up in sequences from peripheral blood control in particular patients. We've ruled out issues of blood being contaminated by somatic tumor tissues, LOH, etc. I gather from the literature that even some canonical EGFR somatic mutations have also been observed in normal tissue controls, and this is now accepted as fact. I would appreciate hearing about your thoughts and/or similar experiences.