As I was browsing through the posters in the Bioinformatics and Genomic Technology section, the title, “To Confirm or Not Confirm” caught my eye. Recalling the quote from Hamlet, I thought the author had missed the second “to” in the phrase. As clarified by Dr. Steven Lincoln, the poster’s author, this wasn’t the case. The poster referred to a specific American College of Medical Genetics and Genomics (ACMG) guideline which states that all NGS results reported back to a physician be confirmed by an orthogonal method such as Sanger sequencing. Dr. Lincoln feels that this step is unnecessary considering the advanced NGS technologies available. He doesn’t suggest stopping the practice completely, though. Instead, he believes each lab should come up with their own strategy to deal with the confirmation of results.
In Dr. Lincoln’s poster, he showed how he used a cross-laboratory framework to develop a strategy and suggested some best practices. The core of his idea is the use of multiple thresholds to decide when a sample needs confirmation by Sanger. Each lab, based on their experience and validation data, should develop criteria to divide their sample results into 3 categories. These categories are samples that are so strongly positive that they need no confirmation, samples that are true positives but require confirmation and samples that are highly likely to be false positives, which should not be included in the report. He believes that the results and framework in the poster will start a conversation on orthogonal confirmation and determine the appropriate burden of proof to potentially change practice.
Another poster also tackled the issue of Sanger sequencing as a validation tool. In “Is Sanger Sequencing Still the Gold Standard?”, Dr. Tyler Beck of NIH said that a single Sanger validation is more likely to erroneously report a false negative than it is to identify false positives. Based on his experience and other experimental data, he concluded that Sanger validation is not necessary. Dr. Beck therefore recommends removing orthogonal Sanger validation of variants from the best practice guidelines.
What do you think about the value of Sanger sequencing as orthogonal validation for NGS? Does it still have value in today’s practice, or are our current NGS methods advanced enough that it’s become obsolete?