Protein-protein interaction (PPI) networks provide a static map of functional protein interactions, which when combined with algorithms, can prioritize key protein candidates which experimental studies cannot capture. This study, aimed to construct knowledge-based nucleus pulposus (NP)-specific PPI networks which could be deployed to investigate complex protein interactions in human NP cells and tissues following IL-4 and IL-10 stimulation. NP-specific PPI networks were developed based on mass spectrometry (MS) and secretome datasets from human NP cells. These networks were validated using in vitro and ex vivo experimental data sets. Genes Underlying Inheritance Linked Disorders (GUILD) genome-wide network-based prioritization framework was employed for protein candidate prediction under no treatment baseline and IL-4, IL-10 and IL-1β single or combined stimulating scenarios. These secretome-based in vitro PPI networks were able to reproduce the no-treatment candidate prioritization baseline. Whereby within NP cells from discs isolated due to traumatic injury biglycan was identified whilst in degenerate samples decorin was highlighted. Furthermore, experimentally observed IL-4 pleiotropic behaviour was predicted by IL-1 receptor-like 1 prioritization. PPI network-based IL-4 and IL-10 conditions offered novel insights of potential candidates, including collagen IV and fibroblast growth factor intracellular binding protein (FIBP) as key candidates within IL-4 activation pathways, whereas urocortin 3 and neural growth factor were identified following IL-10 stimulation. Additionally, MS based PPI network propagation offered a more extensive, module-based structure networks with lower edge degree and biological variability. Overall, multiple proteomic experimental approaches are required to successfully validate in-silico prediction models to understand the complex interactions between the plethora of proteins involved in IVD degeneration.