A machine learning classifier trained on cancer transcriptomes detects NF1 inactivation signal in glioblastoma

Genomic tools allow investigators to devise therapies targeting specific molecular abnormalities in tumors. One such alteration is the loss of neurofibromin 1 (NF1), an important tumor suppressor that regulates the activity of RAS GTPases [1, 2]. Heterozygous mutation or deletion of NF1 causes neurofibromatosis type 1 (NF), one of the most frequently inherited genetic disorders [3]. NF patients often develop plexiform neurofibromas (PNs), benign nerve tumors for which the only therapy is surgery. However, resection is often impossible due to the tumor's intimate association with peripheral and cranial nerves [4].