Development of a generic, structural bioinformatics information management system and its application to variation in foot-and-mouth disease virus proteins

Abstract

Structural biology forms the basis of all functions in an organism from how enzymes work to how a cell is assembled. In silico structural biology has been a rather isolated domain due to the perceived difficulty of working with the tools. This work focused on constructing a web-based Functional Genomics Information Management System (FunGIMS) that will provide biologists access to the most commonly used structural biology tools without the need to learn program or operating specific syntax. The system was designed using a Model-View-Controller architecture which is easy to maintain and expand. It is Python-based with various other technologies incorporated. The specific focus of this work was the Structural module which allows a user to work with protein structures. The database behind the system is based on a modified version of the Macromolecular Structure Database from the EBI. The Structural module provides functionality to explore protein structures at each level of complexity through an easy-to-use interface. The module also provides some analysis tools which allows the user to identify features on a protein sequence as well as to identify unknown protein sequences. Another vital functionality allows the users to build protein models. The user can choose between building models online on downloading a generated script. Similar script generation utilities are provided for mutation modeling and molecular dynamics. A search functionality was also provided which allows the user to search for a keyword in the database. The system was used on three examples in Foot-and-Mouth Disease Virus (FMDV). In the first case, several FMDV proteomes were reannotated and compared to elucidate any functional differences between them. The second case involved the modeling of two FMDV proteins involved in replication, 3C and 3D. Variation between the several different strains were mapped to the structures to understand how variation affects enzymes structure. The last example involved capsid protein stability differences between two subtypes. Models were built and molecular dynamics simulations were run to determine at which protein structure level stability was influenced by the differences between the subtypes. This work provides an important introductory tool for biologists to structural biology.