New database for protein design by artificial intelligence

With the increasing power of supercomputers, artificial intelligence is penetrating more and more areas of scientific life and helping to solve the problem that scientific leaders have been struggling with for decades. A necessary prerequisite for successfully finding a solution is the availability of a sufficient amount of quality source data, based on which the algorithm analyzes the problem. Without data, finding the right solution is much more difficult or impossible. Many scientific communities have not dealt with historically systematic data collection and management, and their current acquisition from thousands of published studies is very time consuming and laborious.

One of the attractive areas for the application of artificial intelligence is the analysis of the effect of mutations on the thermal stability of a protein, as this mechanism is not well understood. This area also suffers from a lack of quality data, and therefore data analysts and programmers from the Loschmidt Laboratories of the Faculty of Science, Masaryk University and the International Clinical Research Center of St. Anne’s University Hospital Brno (FNUSA-ICRC) under the guidance of molecular biologist Mgr. David Bednář Ph.D. and Mathematics Stanislav Mazurenko Ph.D. from the FNUSA-ICRC Protein Engineering research team decided to create a new database FireProtDB, which would systematically collect and maintain this data for a long time.

“The database currently contains 16,000 experimental values obtained from own measurements, available scientific literature and no longer maintained databases, which have been thoroughly filtered and verified,” said Mgr. Jan Štourač, who is one of the creators of the database. To access the data, it offers users a simple web interface, which has been visited by more than 500 scientists from around the world since its publication in October 2020. The data were also provided to the worldwide PDBe-KB database, which serves as a global repository of information for biological and biomedical research and is managed by the European Institute of Bioinformatics. The FireProtDB database was published in the prestigious scientific journal Nucleic Acids Research.

Link to the database here.
.