Using large-scale genomics data to identify driver mutations in lung cancer: methods and challenges

Lung cancer is the commonest cause of cancer death in the world and carries a poor prognosis for most patients. While precision targeting of mutated proteins has given some successes for never- and light-smoking patients, there are no proven targeted therapies for the majority of smokers with the disease. Despite sequencing hundreds of lung cancers, known driver mutations are lacking for a majority of tumors. Distinguishing driver mutations from inconsequential passenger mutations in a given lung tumor is extremely challenging due to the high mutational burden of smoking-related cancers. Here we discuss the methods employed to identify driver mutations from these large datasets. We examine different approaches based on bioinformatics, in silico structural modeling and biological dependency screens and discuss the limitations of these approaches.