Analysis of 10 independent samples provides evidence for association between schizophrenia and a SNP flanking fibroblast growth factor receptor 2.

We and others have previously reported linkage to schizophrenia on chromosome 10q25-q26 but, to date, a susceptibility gene in the region has not been identified. We examined data from 3606 single-nucleotide polymorphisms (SNPs) mapping to 10q25-q26 that had been typed in a genome-wide association study (GWAS) of schizophrenia (479 UK cases/2937 controls). SNPs with P<0.01 (n=40) were genotyped in an additional 163 UK cases and those markers that remained nominally significant at P<0.01 (n=22) were genotyped in replication samples from Ireland, Germany and Bulgaria consisting of a total of 1664 cases with schizophrenia and 3541 controls. Only one SNP, rs17101921, was nominally significant after meta-analyses across the replication samples and this was genotyped in an additional six samples from the United States/Australia, Germany, China, Japan, Israel and Sweden (n=5142 cases/6561 controls). Across all replication samples, the allele at rs17101921 that was associated in the GWAS showed evidence for association independent of the original data (OR 1.17 (95% CI 1.06-1.29), P=0.0009). The SNP maps 85 kb from the nearest gene encoding fibroblast growth factor receptor 2 (FGFR2) making this a potential susceptibility gene for schizophrenia.

A new statistical approach to detecting significant activation in functional MRI.

There are many ways to detect activation patterns in a time series of observations at a single voxel in a functional magnetic resonance imaging study. The critical problem is to estimate the statistical significance, which depends on the estimation of both the magnitude of the response to the stimulus and the serial dependence of the time series and especially on the assumptions made in that estimation. We show that for experimental designs with periodic stimuli, only a few aspects of the serial dependence are important and these can be estimated reliably via nonparametric estimation of the spectral density of the time series, whereas existing techniques are biased by their assumptions. The linear model with (stationary) serially dependent errors can be analyzed entirely in frequency domain, and doing so provides many insights. In particular, we introduce a technique to detect periodic activations and show that it has a distribution theory that enables us to assign significance levels down to 1 in 100,000, levels which are needed when a whole brain image is under consideration. Nonparametric spectral density estimation is shown to be self-calibrating and accurate when compared to several other time-domain approaches. The technique is especially resistant to high frequency artefacts that we have found in some datasets and we demonstrate that time-domain approaches may be sufficiently susceptible to these effects to give misleading results. The method is easily generalized to handle event-related designs. We found it necessary to consider the trends in the time series carefully and use nonlinear filters to remove the trends and robust techniques to remove "spikes." Using this in connection with our techniques allows us to detect activations in clumps of a few (even one) voxel in periodic designs, yet produce essentially no false positive detections at any voxels in null datasets.