BackgroundTo understand the contribution of Mendelian mutations to the burden of undiagnosed diseases that are suspected to be genetic in origin, we developed a next-generation sequencing-based multiplexing assay that encompasses the ~3000 known Mendelian genes. This assay, which we term the Mendeliome, comprises 13 gene panels based on clinical themes, covering the spectrum of pediatric and adult clinical genetic medicine. We explore how these panels compare with clinical whole exome sequencing (WES).ResultsWe tested 2357 patients referred with suspected genetic diagnoses from virtually every medical specialty. A likely causal mutation was identified in 1018 patients, with an overall clinical sensitivity of 43 %, comparing favorably with WES. Furthermore, the cost of clinical-grade WES is high (typically more than 4500 US dollars), whereas the cost of running a sample on one of our panels is around 75–150 US dollars, depending on the panel. Of the “negative” cases, 11 % were subsequently found by WES to harbor a likely causal mutation in a known disease gene (largely in genes identified after the design of our assay), as inferred from a representative sample of 178. Although our study population is enriched for consanguinity, 245 (24 %) of solved cases were autosomal dominant and 35 (4 %) were X-linked, suggesting that our assay is also applicable to outbred populations.ConclusionsDespite missing a significant number of cases, the current version of the Mendeliome assay can account for a large proportion of suspected genetic disorders, and provides significant practical advantages over clinical WES.Electronic supplementary materialThe online version of this article (doi:10.1186/s13059-015-0693-2) contains supplementary material, which is available to authorized users.

No abstract

No abstract

The human androgen receptor (AR) is a ligand-activated transcription factor that plays a pivotal role in the development and progression of prostate cancer (PCa). Many forms of castration-resistant prostate cancer (CRPC) still rely on the AR for survival. Currently used antiandrogens face clinical limitations as drug resistance develops in patients over time since they all target the mutation-prone androgen binding site (ABS), where gain-of-function mutations eventually convert antagonists into agonists. With a significant number of reported distinct mutations located across the ABS, it is imperative to develop a prognostic platform which would equip clinicians with prior knowledge and actionable strategies if cases of previously unreported AR mutations are encountered. The goal of this study is to develop a theoretical approach that can predict such previously unreported AR mutants in response to current treatment options for PCa. The expected drug response by these mutants has been modeled using cheminformatics methodology. The corresponding QSAR pipeline has been created, which extracts key protein-ligand interactions and quantifies them by 4D molecular descriptors. The developed models reported with an accuracy reaching 90% and enable prediction of activation of AR mutants by its native ligand as well as assess whether known antiandrogens will act on them as agonists or antagonists. As a result, a previously uncharacterized mutant, T878G, has been predicted to be activated by the latest antiandrogen enzalutamide, and the corresponding experimental evaluation confirmed this prediction. Overall, the developed cheminformatics pipeline provides useful insights toward understanding the changing genomic landscape of advanced PCa.

No abstract