A genome-wide association study to detect genetic variation for schizophrenia

点击翻译按钮获取中文摘要

DESCRIPTION (provided by applicant): The identification of the specific genetic variants that increase susceptibility to schizophrenia is important to further understand its pathophysiology and develop effective treatments. For this purpose it will be necessary to screen many markers across the whole genome for association with schizophrenia. Such a genome-wide association (GWA) study is currently completed in the CATIE sample and another GWA for schizophrenia will be completed by GAIN in the middle of 2007. However, these current GWAs are perhaps better viewed as "screening" studies and a series of well- designed replication studies may be required to eliminate false positives and validate discovered marker-disease associations. In this application, we will attempt to contribute to the rigorous evaluation of the common disease/common variant model for schizophrenia. Rather than using non- optimal sample sizes and arbitrary rules such as P-values smaller than 0.05 suggest a replication, we will design a series of cost-effective follow up studies using our statistical framework for adaptive multi-stage studies, a systematic bioinformatic approach to integrate WGA data with other sources of information, and a two-day meeting with a panel of experts. The initial replication effort involves US case-control samples (16K SNPs in 5,000 samples with equal cases and controls) where the goal is to detect, with the lowest possible costs, 80 percent of the common genetic variants that increase susceptibility to schizophrenia while controlling the FDR at the 0.1 level. To verify that the identified associations in the case-control samples are not the result of population stratification/ascertainment artifacts and to study possible population differences in the effects of these genes, this is followed by "gene-based replication" study genotyping 300 SNPs in 5,500 individuals from 1,400 families. Finally, instead of merely testing for main effects, we will search for interactions between genotypes and environmental variables plus effects of anti-psychotic medication, use our artificial intelligent "model discovery" software to search for schizophrenia subtypes with different genetic and environment etiology, and search for copy number polymorphisms. All genotype and clinical data will be deposited in the public domain. We attempt to contribute to the rigorous evaluation of the common disease/common genetic variant model for schizophrenia, which is key to further understand its pathophysiology and develop effective treatments.