The purpose of this project is to provide an intuitive web-based interface for the popular bioinformatics program GLIMMER. GLIMMER stands for Gene Locator and Interpolated Markov ModelER and uses interpolated Markov models to identify coding regions in microbial DNA using a reference genome. The web version of Glimmer that is currently available is limited in the options it allows. It also places some restrictions on the files you can use because there is no file upload for one of the sequences. The Glimmer command line software is already widely used in bioinformatics. My goal for this project is to provide a user-friendly web interface for Glimmer for the researchers out there that are not as computer savvy as the rest of us. The front-end of the web app was created using the Foundation 4 package from Zurb. Most of the aesthetic features and JavaScript comes from the package. Perl was used for the back-end along with the Template module for communication between the Perl script and the browser. The entire process of creating the project was carefully planned out from the beginning and thus, there were minimal issues throughout development. There were however, some issues with the CSS on Firefox during browser testing. A workaround had to be written which does not affect the usability of the web app whatsoever. The results of the web app are identical to those you would get by using the command line version of Glimmer 3.02. This is as intended. There is a minor drawback, wherein one cannot set the output of the tool to verbose in order to determine if there is something wrong (debugging). However, this is common with web apps of this kind and the system is designed to display certain errors which can be remedied by the user.
I have noticed that a lot of the scientific software out there is written with only the science in mind. That works well for people with extensive computer experience, who can debug and configure this software. However, to be maximally effective, the software needs to be accessible to those researchers with less advanced computer skills. That is the goal of this project. Almost every option has a tooltip explaining what it is meant to do. The web app also incorporates as much of the original Glimmer documentation as possible. The main search form comes with built-in sample sequences that can be used to test the web app. The overall goal here is to present an easy-to-use web interface for a popular bioinformatics app. There is also the additional advantage of being able to save the results of your query. This is important for the user, but also for the provider of the web app as precious computing time and space does not have to be wasted by running the same queries over and over again. The Perl script is written with this in mind and gets rid of all but one file generated during its analysis. The remaining files could be cleared from the server by scheduling a CRON job to do so.