Proteins are essential parts of all living organisms, and their structure is key to understanding their biological function. In fact, structurally similar proteins (even from different organisms) can have similar functions.

As early as 1971, the Protein Data Bank (PDB) [1] was established as the main repository of protein structures. It currently contains more than 210,000 proteins, adding new structures every week. Recently, the AlphaFold (AF) algorithm has been developed, which can predict the structures of proteins with high accuracy [2]. Currently, there are more than 200 million protein entries in its database.

Many tools exist for pairwise comparison of proteins (e.g., [3]). However, finding similar structures across the entire database, such as PDB or AF, is computationally challenging.

For this reason, in collaboration with the DISA laboratory, a technique for efficiently searching these data was developed, and a prototype web application was created [4]. However, the application is currently limited to the PDB database and does not support automatic updates.

The thesis objectives are:

• To learn about protein structures, protein storage formats and the PDB database.
• To get familiar with the current prototype implementation of the web application (back end, front end, database, computational core).
• Where necessary, to rewrite parts of the prototype implementation to make it easily manageable.
• To add support for automatic updates from the PDB database.
• To explore the possibilities of extending the application with AF data.