Protein-Protein Relationship Measurement Based on MELK Data for Polymyositis

Abstract

Polymyositis is an inflammatory myopathy characterized by muscle invasion of T-cells penetrating the basal lamina and displacing the plasma membrane of normal muscle fibers. In order to understand the different adhesive mechanisms at the T-cell surface, Schubert randomly selected 17 proteins expressed at the T-cell surface and studied them using MELK technique, among which 15 proteins are picked up for further study by us. Two types of functional similarity graphs are constructed for these proteins. The first type is MELK similarity graph, which is constructed based on their MELK data by using the Mutual Information similarity measuring method. The second type is GO similarity graph, which is constructed based on their GO annotation data by using the Maximal Depth method to measuring functional similarity. Then the subset surprisology theory is employed to measure the degree of similarity between two graphs. Our computing results show that these two types of graphs are high related. This conclusion added new values on MELK technique and expanded its applications greatly. Keywords—MELK; Polymyositis; GO; set surprise; mutual imformation