Mainz, Germany, October 2, 2012 – TRON scientists have pioneered a statistical framework that enables researchers to reduce the number of false discovery rates for tumor mutations. The method was published in the September issue of PLOS Computational Biology. While next generation sequencing (NGS) has enabled the highly accurate, unbiased, and high throughput discovery of genetic variations and somatic mutations, the platform is still prone to errors resulting in inaccurate mutation calls. “Even though at 1 to 2 percent the error rate is very low, this will accumulate to thousands of false predictions because of the vast size of the human genome”, says Dr. Martin Löwer, first author and researcher in the TRON Computational Medicine unit. “Improved accuracy is a key prerequisite for the use of new sequencing technologies in a clinical setting.” The authors describe a statistical process that allows for a generic comparison of experimental and computational protocol steps on generated quasi ground truth data. It is applicable for the diagnostic or therapeutic target selection as it is able to discriminate true mutations from false positives, improving on the standard discovery process. The framework, furthermore, enables the definition of best practice procedures for the discovery of somatic mutations.
Löwer, M., Renard, B.Y., de Graaf, J., Wagner, M., Paret, C., Kneip, C., Türeci, Ö., Diken, M., Britten, C.M., Kreiter, S., Koslowski, M., Castle, J.C., Sahin, U. (2012) Confidence-based Somatic Mutation Evaluation and Prioritization. PLoS Comput Biol 8(9): e1002714.