Testing one two.

When you order germline genetic testing for cancer patients, how often do you think about the accuracy of that test? Genome Analysis Toolkit is a standard tool used to sift through raw genomic data and identify pathogenic mutations. In this this study, a deep learning algorithm called DeepVariant was compared to the standard Genome Analysis Toolkit using two large cohorts of patients with prostate cancer and melanoma. Each method agreed on about 92% of pathogenic variants while others were detected by one method but not the other. Overall, the deep learning algorithm performed better and accurately detected more variants than the standard approach from the same raw data. But what struck us when reading this study was how much we take for granted when it comes to genetic testing. This study points out the potential variation in accuracy and quality of germline genetic testing--something that’s not heavily regulated in the U.S. TBL: Just because a patient has a pathogenic variant in a cancer predisposition gene doesn’t mean it will be identified by germline testing. It is critical to understand the importance of test and laboratory quality in accurately generating test results. | AlDubayan, JAMA 2020