Large-scale proteomic methods can identify quantitative differences in proteome-wide protein levels between control and diseased neural cells.14 In Figure 2, we show that pairwise SILAC (stable isotope labeling by amino acids in cell culture) comparisons of control and SZ hiPSC NPCs identified changes in cellular adhesion and oxidative stress pathways. SILAC (Supplementary Figure 5A), in conjunction with multidimensional protein identification technology, was used to identify proteins that were up- or downregulated in four pairwise comparisons (consisting of 5 ‘heavy' and 5 ‘light' quantitative mass spectrometry analysis runs per cell line) between gender-matched SZ (P1 and P3) and control (C1, C3 and C6) hiPSC NPCs. Although labeling efficiency was high (Supplementary Figure 5C), ratio-of-ratios analysis (control-heavy/SZ light relative to control-heavy/control light) permitted correction of any incompletely labeled proteins by the fractional analysis. When the same 74 markers of neural subtype identity used to compare microarray gene expression profiles were considered across our four SILAC protein data sets, only one (NCAM1) was significantly perturbed in both patients analyzed (Supplementary Table 2). Through four independent SILAC experiments, we identified perturbed proteins with corrected
