Simultaneous Identification of Rare and Common Segment Variants

• Activity Number: 468
• Type: Contributed
• Date/Time: Wednesday, August 1, 2012 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): X. Jessie Jeng*+ and Tony Cai and Hongzhe Li
• Companies: University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
• Keywords: Recurrent genetic signals ; Optimal adaptivity ; CNV analysis

Abstract

The identification of recurrent signals based on a large set of samples is one of the central topics in population-scale genomic data analysis. A bottleneck of recurrent signal identification is the lack of an adaptive information pooling procedure which can automatically adjust to the unknown signal carrier's proportions and efficiently identify both rare and common recurrent signals. We specifically consider the identification of recurrent DNA copy number variants (CNVs) based on a large set of samples from Neuroblastoma patients. It is likely that both rare and common CNVs cooperate to increase the risk of the disease. We developed the Proportion Adaptive Segment Selection (PASS) procedure, which can optimally and simultaneously identify both rare and common CNVs. The proposed method is statistically rigorous and computationally efficient. Theory, simulation, and real applications will be presented to demonstrate the performance of the method.