Description of datasets.

For this study, three datasets were used; schizophrenia cases and controls, an antipsychotic-free set and a control dataset. Age and gender information is given for cases and controls separately. Gene expression data was generated in two batches (batch 1: Illumina H-8 and batch 2: Illumina H-12) and collected at different sites, information given in the fourth and fifth row). The batch effect resulting from the use of different arrays on different time points in the latter set was removed using the SampleNetwork R package [62]. The number of expressed genes is given in the last row. *DK  =  Denmark and NL  =  The Netherlands.