Here, our set-based analyses provide additional evidence for association between allelic variation in subunits of the SWI/SNF chromatin remodeling complex and AD in the COGA adult case-control sample, and with behavior problems in a younger sample of adolescents. We have now observed association between multiple, different, members of the SWI/SNF complex and AD, as well as a genetically related developmental precursor to AD, in three independent human populations. As hypothesized, secondary analyses of the individual genes in the set indicated that it was not always the same genes contributing to the association signal observed with the complex. In the IASPSAD sample, SWI/SNF complex members SMARCA2 and BRD7 were associated with AD (Mathies et al., 2015). In COGA, we found that the strongest association signals came from SS18L1, SMARCD1, BRD7, SMARCB1 and BCL11A. When we examined the Spit for Science sample, we found that the strongest signals came from ACTB, ARID2, BCL11A, BCL11B, BCL7B, BCL7C, DPF2, and DPF3. This combination of data provides growing support for the involvement of SWI/SNF chromatin remodeling in the development of AD. Importantly, our analyses indicate
