Structural Insights into the Regulation of a Key Tumor Suppressor

Three variants of the confidence set inference (CSI) procedure were proposed and applied to both the simulated and the Collaborative Study on the Genetics of Alcoholism (COGA) data. For each of the two applications, we first performed a preliminary genome scan study based on the microsatellite markers using the GENEHUNTER+ software to identify regions that potentially harbor disease loci. For each such region, we estimated the sibling identity-by-descent sharing probability distribution at the putative disease locus. Based on these estimated probabilities, the CSI procedures were employed to further localize the disease loci using the single-nucleotide polymorphism markers, leading to confidence intervals/regions for their locations. For our analysis with the simulated data, we had knowledge of the simulating models at the time we performed the analysis. Background A frequently used strategy in linkage analysis is to first screen the entire genome using microsatellite (MS) markers, and then to follow up on preliminary linkage regions using densely saturated (often single-nucleotide polymorphisms (SNP)) markers. Although many statistical methods are available for each step of this two-stage approach, only a limited number of them (e.g., Liang et al. [1]) are able to provide confidence estimates of disease gene locations. Furthermore, most of the methods available are subject to multiplicity adjustment, which is a non-trivial matter given the complex dependency of the statistics involved. The confidence set inference (CSI) procedure [2] can be used to obtain confidence estimates using affected sibpair (ASP) data, and avoids the multiple testing problem. Unlike the approach of Liang et al. [1], it is not based on the asymptotic distribution of the estimator of the location of the trait locus. Instead, it indirectly deduces a confidence region for the trait locus based upon a set of markers that are inferred to be within a pre-specified distance from the trait locus. Note that this is a non-directional procedure that makes no distinction between loci symmetrically located around a marker. In the present paper, we explore three variants of the CSI procedure to further improve its performance. The first modification is to test every location (not just the markers) in the region of interest. This practice effectively eliminates the "nondirectional" problem of the original CSI method. The second variant is a multipoint extension of the first variant, in that information from all markers are utilized to calculate the IBD sharing statistic at a marker locus. The last is also a multipoint approach, but the identity-by-descent (IBD) sharing statistic is calculated at each hypothesized disease locus rather than at its nearest marker locus. from Genetic Analysis Workshop 14: Microsatellite and single-nucleotide polymorphism Noordwijkerhout, The Netherlands, 7-10 September 2004 Published: 30 December 2005 BMC Genetics 2005, 6(Suppl 1):S21 doi:10.1186/1471-2156-6-S1-S21 Genetic Analysis Workshop 14: Microsatellite and single-nucleotide polymorphism Joan E Bailey-Wilson, Laura Almasy, Mariza de Andrade, Julia Bailey, Heike Bickeböller, Heather J Cordell, E Warwick Daw, Lynn Goldin, Ellen L Goode, Courtney GrayMcGuire, Wayne H ning, ail Jarvik, Brion S Maher, Nancy Mendell, Andrew D Paterson, John Rice, Glen Satten, Brian Suar z, Veronica Vieland, Marsha Wilcox, Heping Zhang, Andre s Ziegler and Jean W MacCluer Proceedings