BUILDING AND MINING GRAPH DATABASES FROM BIOMEDICAL HETEROGENEOUS NETWORKS
All things and phenomena in life, especially in the fields of life and biomedical medicine, are more or less related, interact with each other, forming a heterogeneous network. Therefore, when studying an object, we need to consider the relationships around it. However, current research often focuses on a specific object, not considering other subjects that are influencing it. Therefore, this paper proposes to use a graph database as an approach to dealing with heterogeneous networks solving the biomedical problem. Experimental results on two heterogeneous networks, miRNA-disease, and autism-miRNA-protein, has drawn the network of interactions, the relationships in a very intuitive way; shows the interaction between each specific object in the graph; and finally, statistics the interaction levels and shows the top 5 diseases, the top 5 miRNAs with the most interaction in the data. From there, it can be seen that the proposed method improves efficiency, increases accuracy, and reduces execution time compared to the traditional way of storing data before.