Towards accurate detection of somatic variations in contaminated tumor samples

Abstract

Somatic substitutions are the most frequent type of mutations that are known to drive most human cancers. It has been key interest of the cancer genomicists and clinicians to reliably and accurately detect such somatic substitutions in tumor samples. Major efforts have been used to develop such tools for the past several years and as a result, many software tools have been reported. However, it is still difficult to obtain pure tumor samples from most cancer patients, a problem that continues to remain as a major challenge for all the existing mutation detection tools. In this talk, we will briefly discuss about a benchmark result for the currently available software tools for detecting somatic substitutions and about our recent efforts to improve the detection accuracy of somatic variations in contaminated tumor samples.