Global proteomics analyses are traditionally performed in data-dependent acquisition (DDA) mode, which results in inadequate reproducibility across large sample cohorts due to the under-sampling inherent to shotgun proteomics. Recently, data-independent acquisition (DIA) strategies were introduced to allow reproducible detection and quantification of thousands of proteins with consistent sensitivity across samples. Here, we present an approach to analyze changes to the protein network in human peripheral blood mononuclear cells (PBMCs) from clinical blood samples, using DIA as a unique platform for biomarker discovery. We describe how to generate spectral PBMC proteome libraries by applying peptide fractionation followed by DDA analysis, and then how to apply DIA methods to PBMC samples from individual patients using a high-resolution Orbitrap Fusion mass spectrometer.