In bioinformatics, graphs are often used to describe cell processes, representing interactions between proteins in a protein-protein interaction network. Some proteins significantly influence these tissues and play an essential role in a regulation called significant proteins. Significant proteins are functional as data for drug discovery. Graph analysis is needed to reduce the graph and find these significant proteins. So far, there have been no studies discussing protein interactions in avian influenza type A/H9N2, a zoonotic subtype of avian influenza. Although this virus is classified as Low Pathogenic Avian Influenza, it is troubling the public because it reduces the production of chicken eggs by 80% and disrupts the reproductive organs of mammals. This study aimed to find a significant protein from avian influenza type A/H9N2. From the results of this study, we found ten network clusters with the ClusterONE algorithm. We perform graph analysis on the first cluster because it is the best cluster with the smallest p-value of 0.000023. The cluster contains 20 nodes representing 20 proteins. The analysis graph (Analysis centrality) that is simulated in MATLAB includes betweenness centrality, closeness centrality, degree centrality, eigenvector centrality, and page rank centrality. Of the 20 proteins, nine significant proteins were obtained, namely CD247, CD48, FCGR2A, FCGR2B, IL10, PTPN22, PTPRC, SLAMF1, and TRL5, the highest score on the FCGR2A protein. The highest score indicates that the FCGR2A protein has the most dominant effect on the protein-protein interaction network of Avian Influenza virus type A/H9N2.