A new visualization tool, called "Visual Concept Explorer (VCE)", was developed to visualize concept relationships in bio-medical literature. VCE integrates pathfinder network scaling and Kohonen self-organizing feature map algorithm for visual mapping. As a case study, VCE was applied to visualize a chromatin protein-protein interaction (PPI) network. The mapping results demonstrated that VCE could explore the semantic structure and latent domain knowledge hidden in protein-protein interaction data sets generated from bio-medical literature.