@article{article_1737250, title={Integrative Network-Based Characterization of Serum Biomarker Interactions in Prostate Cancer Progression}, journal={Güncel Tıbbi Araştırmaları Dergisi}, volume={5}, pages={7–18}, year={2025}, author={Erbak Yılmaz, Huriye and Özel, Zeynep and Dinçkal, Çiğdem and Karalar, Oğuz and Güç, Zeynep Gülsüm and Dinçkal, Mustafa and Tekindal, Mustafa Agah and Şentürk, Şerif}, keywords={Prostat kanseri, Serum biyobelirteçleri, Ağ analizi, Trombospondin-1, Nöropilin-1, Hipoksiye duyarlı faktör-1 alfa}, abstract={Objective: This study investigates the interactions among key serum biomarkers—Hypoxia-inducible factor-1 alpha (HIF-1α), Thrombospondin-1 (TSP1), Neuropilin-1 (NRP1), and Prostate-specific antigen (PSA)—and clinical parameters including age at diagnosis, diabetes mellitus (DM), hypertension (HT), and smoking status in patients with metastatic prostate cancer (mPCa). The objective was to identify structural and functional interdependencies among these variables using a network-based analytical approach.
Materials and Methods: Network analysis was conducted using JASP software (v0.19.3.0). Variables were modeled as nodes, and partial correlations between them as edges. Edge color represented the direction (positive or negative) of the correlation, while thickness indicated its strength. Network topology was evaluated using graph-theoretical metrics including degree, closeness, betweenness, and eigenvector centrality. Additional measures of density and sparsity were also calculated. Spatial visualization of the network was performed using the Fruchterman–Reingold algorithm.
Results: The network comprised eight variables and 27 connections, yielding a sparsity value of 0.036, indicating a highly dense structure. PSA and TSP1 exhibited the highest betweenness centrality, serving as critical bridging nodes. HT and DM had high degree and closeness centrality values, reflecting central positions within the network. NRP1 displayed the highest clustering coefficient, suggesting a localized regulatory role. A strong negative association was observed between TSP1 and HT.
Conclusion: This study highlights the utility of network analysis as a systems-level tool to explore complex biomarker interactions in mPCa. PSA, TSP1, and NRP1 emerged as key molecular regulators, while systemic conditions such as HT and DM significantly influenced network architecture. These findings warrant further validation through mechanistic and hypothesis-driven statistical studies.}, number={2}, publisher={İzmir Katip Çelebi Üniversitesi}