期刊名称：International Journal of Computer Information Systems and Industrial Management Applications
印刷版ISSN：2150-7988
电子版ISSN：2150-7988
出版年度：2009
卷号：1
页码：321-328
出版社：Machine Intelligence Research Labs (MIR Labs)
摘要：A plethora of online structure prediction for proteins servers can be found over the internet. Increasing complexities during structure prediction may be attributed to the quotidian rise of these servers. One of the key limitations of these servers is the lack of rapid, robust modeling, pa ltry results and the inability to give reproducible methods for detecting, matching and analyzing protein structures. The most commonly used approaches involve getting an unknown sequence (and a template, in some servers) and then aligning the sequence with its own pre-defined parameters, then obtaining a predicted secondary structure. We introduce a server, using Java Server Pages architecture upon Apache Tomcat, for protein structure prediction wherein the user is provided maximum control over the parameters and variables, which define the relationship and homology of the unknown sequence with known sequence databases. Added to that, the package uses Profile Hidden Markov Models for template selection and Python programs of Modeller for structure prediction based on the selected templates. It coherently implies the user doing his modeling studies as if on (more reliable) standalone software, and that too, without the hassles of any coding, writing tedious scripts or fallacious guess work resulting in protracted homology models. There is an auxiliary module to the server to determine the composition of the user's unknown protein structure and correct mistakes, if any. All the steps in course are fully transparent so as to give full independence of changing the variables as and when, suited to the user, to get perfect results.
