3D structural modeling of HIV-1 protein (protease)

• Main Author: Liu, Xinyi.
• Other Authors: Kwoh Chee Keong
• Format: Final Year Project
• Language: English
• Published: 2011
• Subjects: DRNTU::Engineering::Computer science and engineering::Computer applications::Life and medical sciences
• Online Access: http://hdl.handle.net/10356/46420
• Institution: Nanyang Technological University

The human genome itself operates like a factory. Our body is made up of DNAs, RNAs and mostly proteins. Proteins are vital for our survival, and provide various functions in the human living cells. The way a protein functions is determined strictly by its structure, which is formed by a sequence of amino acid residues. But with constant exposure to various diseases and virus, mutations of the protein may occur; which some may be disastrous to the human body. Hence, it is important to model protein structure to aid in better understanding how it functions. With technology advances, there are tools available to model protein structure. In this project, two available tools, I-TASSER and MODELLER will be evaluated as to which can contribute a more accurate protein structure prediction. HIV-1 aspartyl Protease is used as the target in this project. The tasks involved will investigate the modeling of 3D protein structures by providing the amino acid sequences into both the I-TASSER server and MODELLER tool. Six mutations of protease will be studied to analyze the effects on the protein structures, by comparing how much the mutated structures deviates from the original protease. After building the structural models, PyMOL will be used to align the structures to generate the respective RMSD values. Next, structural minimization energies will be calculated using Swiss-PdbViewer. And eventually, the results which consist of the structures modeled, RMSD and minimization energies will be displayed in a web interface. The eventual results will lead to a conclusion that MODELLER is a more accurate tool for the modeling of protease. The structures modeled are more stable and more consistent as compared to I-TASSER. The size of protease is relatively small of 198 amino acid residues in length, therefore the chances of finding templates that covers a wider range of the amino acid sequences is higher in MODELLER. The structural alignment between mutated proteases and original protease did not affect much the structural changes of the protease structure formation. However, by going more in-depth into looking at what amino acids are involved in forming the sequences, the difference in mutations can be identified. It can be seen that the vast difference in RMSD and minimization energies are contributed by the presence of Phenylalanine and Methionine. This is reflected in the comparison between the most and the least similar structures when alignment with original protease is performed. Results also reflect that mutated protease 6 gives the best evaluation results among other mutated sequences. Thus, it may be considered being use as a drug-target in future research.