The thesis aims to design and implement a web backend application that will, using a REST interface, provide an easy submission of bioinformatics calculations on a computing cluster for web tools developed in the Loschmidt Laboratories at the Faculty of Science, MU. The backend must provide uniform interfaces for working with virtual sessions, files, computational tasks, as well as parsers of output files of bioinformatics applications into JSON format. The backend must implement authentication of multiple users with variable priviledges. Provided APIs for running computations and parsers must be as generic as possible, allowing for the easy addition of new types. All APIs, calculation rules, and parsers must be easily versionable with backward compatibility support so that when a new version is released, all clients need not be updated. The basic and most crucial non-functional requirement is the robustness of the solution so that the server can handle recovery after outages, migrations to newer versions, and try to protect clients from technical (i.e., not caused by user data) problems on the computing infrastructure or third-party services. A client library in the Java programming language must be developed for the backend services.