Graph Analysis for the Discovery of Key Proteins in Type 2 Diabetes Mellitus

Abstract

One of the metabolic diseases with a rising prevalence in Indonesia is Type 2 Diabetes Mellitus (T2DM). A collective effort from various sectors is required to seek solutions for T2DM. The proteomic approach, which focuses on proteins and their interactions related to T2DM, can be used to understand this condition. This research aims to model protein interactions associated with T2DM using a network graph, enabling the identification of key proteins that have the potential to serve as therapeutic targets or T2DM biomarkers. The study begins by acquiring data on T2DM-related proteins and their interactions. Subsequently, data preparation is performed to construct an interaction network graph. The graph is then analyzed using four centrality measures: degree centrality, closeness centrality, betweenness centrality, and eigenvector centrality, to identify key proteins. Finally, gene set enrichment analysis (GSEA) is conducted on the identified key proteins. The analysis of the interaction network graph successfully identifies key proteins associated with T2DM. Out of 27 T2DM-related proteins, seven are selected by all four centrality measures: ABCC8, HNF4A, INS, KCNJ11, NEUROD1, PDX1, and SLC30A8. CAPN10, IL6, and WFS1 are chosen by all centrality measures except for betweenness centrality, while PTRN is only selected by betweenness and eigenvector centrality. GSEA results also indicate that these key proteins play roles in various biological mechanisms related to T2DM. This research successfully identifies T2DM key proteins using graph analysis, potentially serving as therapeutic targets or T2DM biomarkers. However, medical validation of the identified key proteins is necessary.