Due to the cellular complexity of brain tissue, gene expression analysis often result in only modest fold-changes in expression, which further complicates establishing criteria for what constitutes a significant change. Significance has been defined in a number of different ways, such as arbitrarily preset cut-offs using fold-change17–18 or statistical significance19–24 criteria or a combination thereof.25–26 Most microarray studies of substance use employing human postmortem brain report hierarchical lists of “significantly changed” transcripts for a majority of individual case-control studies19–23 or for pooled groups of index and of control cases.17–18,24–25,27 The study design and manner of comparative analysis dictates what constitutes relevant change for a specific microarray platform. Although the use of different platforms, chemistries and data processing hampers direct comparisons between studies28, it is interesting to note that some individual transcripts and functional groups are repeatedly encountered across studies, substances, brain region and/or experimental cohorts (Table 1).
