other SNPs with the same properties, and 2) “cluster then converge”, based on SNPs within genes that are identified in each of the two samples by clusters of SNPs that a) display nominally significant case vs control allele frequency differences and b) lie near other SNPs with the same properties. Since approach (2) does not require that the identical SNPs display nominally significant results in each of several samples, it is especially useful for evaluating concordance between GWA datasets that use different sets of SNPs. We can thus apply this approach to examining the concordance between addiction and co-occurring traits likely to display complex genetic influences [25] using empirical Monte Carlo statistics to assess the significance of results. We discuss this work in light of its technical and analytic limitations and in its similarities and differences with “template” GWA analyses that seek associations that display genome-wide significance, typically in phenotypes that display oligogenic genetic architectures and/or in larger samples that are often recruited in multiple locations. We also describe the ways in which these pooled genotype data identify a number of the same genomic regions that are identified by recently available dbGAP datasets that provide individual genotyping for cocaine-dependent and
