A New Procedure to Identify Neighborhoods of Adjacent SNPs That Show a Specific Biologically Meaningful Pattern of Associations with Multiple Clinical Endpoints

• Activity Number: 609
• Type: Topic Contributed
• Date/Time: Thursday, August 5, 2010 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Stan Pounds*+
• Companies: St. Jude Children's Research Hospital
• Keywords: change-point ; genome-wide association studies ; integrated analysis of genomic and clinical data ; incorporating prior biological knowledge ; projection onto the most interesting statisitcal evidence

Abstract

The change-points in interesting pattern statistics (CHIPS) procedure is proposed as a method to identify segments of adjacent SNPs that exhibit a prespecified biologically interesting pattern of association with multiple clinical outcome variables. Prior biological knowledge is used to define a test statistic that measures evidence for the specific assosiciation pattern of biological interest. These statistics are computed for each SNP. A change-point algorithm is applied to the series of test statistics to segment the genome into regions with SNPs of similar levels of statistical significance. Finally, statistical significance is determined by repeating the entire procedure within a permutation assesment. The CHIPS procedures improves statistical power by leveraging prior knowledge about the relationships among clinical variables and pooling information across neighboring SNPs.