A statistical model to find disease genes Genetic variation is high among individuals, which makes it difficult to identify any one specific pathogenetic variant in patients with idiopathic disease, especially those that are in noncoding regions of the genome. Examining tissue-specific and population-level RNA sequencing data, Mohammadi et al. developed a statistical test, analysis of expression variation (ANEVA), that can quantify how one individual's gene expression fits in the context of the variation within the general population. By applying ANEVA to a dosage outlier test, the authors identified pathogenic gene transcripts in patients with Mendelian muscle dystrophy. Science , this issue p. 351