Comparative Modeling, Characterization and Energy Minimization Studies of Aquaporin 9: An Exclusive Target Protein for Rheumatoid Arthritis
Background: In the absence of the experimentally determined structure, computer aided protein structure prediction, evaluation and their energetically stable structure identification is the only way out of the problem. The main objective of the study was to perform the structure prediction of Aquaporin 9 (APQ9) the most targeted protein for rheumatoid arthritis, using in-silico methodology and validate the generated models. Methods: Secondary structure prediction of AQP9 was performed using GOR4, SOPMA and CFSSP algorithm. This was followed by the three-dimensional structure identification from MODELLER, LOMETS and MUSTER server. Many models were built and the best amongst them was identified on the basis of their DOPE score. RAMPAGE was used to validate these models and finally the selected model was energetically stabilized. Results: Amongst the 4 predicted models, model predicted using MODELLER software with 1FX8 PDB template (MODELER MODEL 2) was selected as the best. This model showed the best results in Ramachandran plot validation. In the Ramachandran Plot, 223 residues (95.7%) were found to be in the favored region, 9 residues (3.9%) in the Allowed region and the rest 1 residue (0.4%) in the Outlier region. Energy minimization calculations were also done for the four models using SPBDV software and Modeler Model 2 model showed the least energy (E= 3484.038 KJ/mol). Conclusion: The accurate three-dimensional structure prediction of proteins is a grand challenge now. Massive amount of sequence and structural data is available now with low cost. The choice of one or other method depends solely on the type of protein sequence and the quality of the predicted structure. The accurate structure prediction, fold recognition, energy calculation, side chain modeling and target template identification are the crucial edges of the molecular modeling process which need to be scrutinized for the best predicted model.