Correlation of Diabetes Mellitus and Cellular Components using Fuzzy K-Partite

Abstract

PPI clustering is one of the computational methods to identify proteins that affect type 2 diabetes mellitus. One of the graph-based fuzzy clustering algorithms, namely fuzzy k-partite clustering was built to solve the problem of biological network data that has more than one function and is present in more than one cluster. which may have overlapping cluster members. A previous study analyzed the mechanism of herbal medicine using the fuzzy k-partite graph clustering method, it was found that there are three groups of proteins that have the same role in overcoming type 2 DM. in the form of backbone tissue (GO-protein). The stages in this research are type 2 DM protein data, search for significant MCL clustering proteins, mining of cellular components in the Uniprot web database, adjacency matrix construction, bipartite network formation with Fuzzy k-Partite Clustering and cluster analysis. This shows that the output of using the algorithm is a network that can provide information on biological processes in type 2 DM. If the weight of the relationship between clusters is high, it can be ascertained that the value of the degree of membership in the cluster is low and there are few cluster members. Conversely, if the weight of the relationship between clusters is low, the degree of cluster membership is high and there are many cluster members. In other words, the value of the degree of membership resulting from the application of this algorithm is inversely proportional to the value of the connectivity between clusters.