LAUSR

In this issue of EBioMedicine, Gatza and colleagues show through an integrative proteogenomic analysis of publicly available datasets that amplification and overexpression of ADNP (Activity Dependent Neuroprotector Homeobox) are associated with poor prognosis in high-grade serous carcinoma (HGSC), and further demonstrate its oncogenic potential via in vitro studies [1]. HGSC is the most common histologic subtype of ovarian carcinoma typically presenting at advanced stage and accounting for most deaths from ovarian cancer [2]. Nearly all HGSCs exhibit TP53 mutations, DNA copy number alterations and multiple subchromosomal aberrations, most frequently MYC, PIK3CA, AKT1 and NOTCH3 amplification, and dysregulation of Rb/E2F, Ras/PI3K and FoxM1 pathways [3,4]. Unfortunately, inhibition of these pathways has not been effective in clinical trials [5], and there are conflicting results for the prognostic significance of molecular subtypes of HGSC [3,6,7]. Therefore, identification of novel oncogenes such as ADNP is essential in the discovery of much needed new prognostic and predictive markers in HGSC. The Gatza group applied the Poor Prognosis Signature (PPS) by the Cancer Genome Atlas (TCGA) Project [3] to three independent datasets [3,7,8] to identify the top 10 signaling pathways associated with the PPS signature. The authors used a 3-step approach to identify functionally significant alterations. First, samples were scored to detect copy number changes associated with a high PPS score. Secondly, the correlation between mRNA or protein expression and the PPS score was assessed to determine the functional relevance of genes/proteins in each amplicon. Finally, the statistically significant candidates were compared to identify genes amplified and overexpressed at the mRNA and protein level. Of 131 genes, 39 were reproducibly correlated with the PPS score. Finally, to prioritize these 39 genes for functional studies, the Achilles RNAi proliferation screen data was assessed. By examining the essentiality of the »9000 genes