Figure 3
Case-specific platelet membrane protein pooling and population results from quantitative proteomics. (A) Five control (C) and five probable Alzheimer's disease (AD) case membrane proteomes that made up the control and AD platelet membrane pools were visualized by silver staining to confirm equal contributions to the pool and to demonstrate comparable purity and integrity. (B) Gaussian fits of the null experiment (red curve) and experimental comparison replicate one (blue curve) for the population of 1,009 quantified proteins binned according to log2-transformed ratio quantified relative abundance. Significance cutoff at ±1.17 is shown as dashed vertical lines. (C) False positive counts (red trace), candidate marker proteins (blue trace), and effect on false discovery rate (FDR) (dashed green trace, scale to right) of various filtering criteria. FDR was 6.9% at the selected criteria of 1.17 log2-transformed AD/CT ratio (average of two technical replicates), coefficient of variance <100%, and signal-to-noise minimum control level greater than 10.