Applied bioinformatic and statistical approaches to complex disorder gene mapping.

摘要

There has been great success in identifying the molecular variants that give rise to Mendelian disorders such as cystic fibrosis and Huntingtons disease. However, the same approaches have not been as successful when applied to complex disorders. A number of explanations and solutions have been proposed. Current challenges include but are not limited to; (1) the linkage regions are broad and contain many genes, (2) these same linkage results do not satisfy proposed criteria for statistical significance, and (3) the basic genetic architecture is still not understood. Schizophrenia is as an example of a complex disorder and was the phenotype used to attempt to address each of these issues. First, bioinformatics and data integration of a variety of types of evidence was used to prioritize candidate genes and polymorphisms in a large linked region. Second, the significance of linkage results was determined empirically. Finally, knowledge of a proposed schizophrenia risk variant was used to clarify linkage evidence in another region of the genome. The results of this research show; (1) bioinformatics tools can be used to rapidly prioritize genes in a large linked region, (2) the NPL Z-score is not distributed as a standard normal, (3) stratification of families based on a proposed DTNBP1 risk haplotype on chromosome 6 increases the evidence for linkage on chromosome 8, and (4) the elusive goal of genome-wide significant linkage evidence is demonstrated via simulation. Each of these steps has advanced the goal of identifying additional schizophrenia susceptibility loci and understanding the genetic architecture of a socially and economically important disorder. Further, the approaches are general and can be applied to any complex disorder or trait.