Munazzah Tasleem1, B Jayaram2
1Department of Computer Sciences, Jamia Millia Islamia, Delhi, India.
2Indian Institute of Technology, Delhi, India.
ABSTRACT
To assess the accuracy of Bhageerath_H and to suggest improvement in it so as to take it to perfection where it gives the best modeled structure very close to the native with RMSD below 5Å.The aim of Homology Modeling is to build protein model close to the native with RMSD less than 5Å, even when the Template identity is less than 30%. Bhageerath_H gives five models for every query submitted. Bhageerath_H operates in three modes: Auto Template Search Mode, Exclude Native Mode, and User Defined Mode. We modeled structures of CASP3, CASP7, CASP8 and CASP9 target sequences using Bhageerath_H in Auto Template Search mode, and compared the modeled structures with the native structures using Pymol and Superimpose. We modeled structures of CASP8 target sequences using other homology modeling softwares and compare the modeled structure with the native structure and observed that Bhageerath_H is building models as good as any other homology modeling tool when the sequence similarity is above 90%. We modeled CASP9 target sequence structure using Bhageerath_H in exclude native mode and compared the structure. We collected data for Protein coding genes of Plasmodium falciparum using NCBI and PlasmoDb. All the protein sequences which are experimentally determined are identified and the similarity of all the protein sequences of P. falciparum with human proteome is identified using BLASTp. The protein sequences with less than 50% similarity with human proteome were extracted and the possible drug targets were identified. 15 possible drug target sequences were selected for homology modeling and their structures were modeled using Bhageerath_H, and validated the modeled structures.Validation of Bhageerath_H is required to model the native like structure of possible drug target sequences of P. falciparum, so as to reach at the states where the drug to cure Malaria is identified computationally. Read More …
