Type 2 diabetes mellitus (T2DM) is the most common type of diabetes in the world. The second line of T2DM treatment that can inhibit glucagon release and increase insulin secretion is DPP-IV inhibitor. The DPP-IV enzyme in human body can be produced by the body itself or bacteria found in gut, such as Parabacteroides goldsteinii and Bacteroides fragilis. In this study, bacterial sequence alignment and homology modeling were carried out. Molecular docking of human DPP-IV and bacteria model was carried out on each active site of human DPP-IV using PyRx. Percentage of sequence similarity between human DPP-IV and bacterial models (%conserved sequence;%overall sequence) is P.goldsteinii (46.15%,24.18%) and B.fragilis(92.31%;20.04 %). Optimization parameter for molecular docking is gridbox 50x50x50 units with spacing of 0.375 and energy evaluation of 5.000.000. Based on analysis result of selectivity index values, obtained DPP-IV inhibitor, namely gemigliptin that relatively selective in P.goldsteinii with selectivity index value of 1.34x10-4 and retagliptin that relatively selective in B.fragilis with selectivity index value of 0.05x10-4 between 16 ligands in active site of DPP-IV enzyme. This showed that gemigliptin and retagliptin can be used as supporting data in new molecular design and their development as selective DPP-IV inhibitor for T2DM patients.

Keywords: Bacteroides fragilis; DPP-IV inhibitor;homology modelling; Parabacteroides goldsteinii;selectivity