The major limitation of most genomic studies is the focus on individual genes with the highest statistical significance. In this study we utilized a network based approach, to establish a better functional understanding of changes occurring in the NAc transcriptome of subjects with AD. All nominally significant transcripts identified in the univariate analyses (p≤ 0.05) were used to generate AD-relevant gene co-expression networks. This significance threshold was chosen to allow the inclusion of true positive signals with smaller effect size (which would otherwise be excluded with more stringent statistical criteria) and to retain a sufficient number of genes with biological importance in AD in the building of gene co-expression networks. A total of 24 modules were identified, including the grey module (M grey), which contains 13 transcripts unassigned to any of the other 23 modules [41]. The module sizes varied from 1106 transcripts in M turquoise to 35 transcripts in M darkgrey (Fig 1A). Next, to assess the quality of sample matching and to detect any confounded modules, the module eigengenes (MEs), which represent the sum of gene expression profiles
