Predikce proteinových komplexů

Abstract

The aim of this thesis was to develop an application to support experiments focused on the prediction of protein complexes. Protein-protein interactions are fundamental biochemical processes that enable cellular cooperation through physical contacts between proteins. These interactions form protein complexes which are essential for numerous biological functions. Predicting such complexes is a challenging task that requires advanced algorithms, high-quality datasets, and sophisticated analytical tools.

The proposed application combines modern web technologies. The front-end is built using TypeScript and React to provide an interactive and user-friendly interface, while the back-end is implemented in Python as an API responsible for data processing and communication with protein interaction databases. The tool supports not only the prediction of protein complexes but also their analysis and quality evaluation.

The thesis is structured into eight chapters, starting with a theoretical introduction to protein-protein interactions, followed by an overview of selected prediction algorithms and the technologies used, and concluding with implementation, testing, and evaluation of the developed application.