The Austrian Centre of Industrial Biotechnology (acib) and the University of Graz have teamed up to develop a combination of database and search engine that can filter enzyme functions out of the thousands of protein structures used in the chemical, cosmetic and pharmaceutical industries.
The Catalophor System has been designed to filter results and can even track functionalities that have not been discovered yet.
Enzymes are gaining importance as environmentally friendly, important components of industrial processes in the chemical industry, and according to acib, the search for industrial usable enzyme functions was previously extremely complex and subjected to restrictions.
This new search engine, including database with more than 100,000 proteins opens up new possibilities in the search for new enzyme functions.
The procedure is similar to a Google search, although the input of data is a bit more complicated. It all starts with the question for the required enzyme function.
“We focus on the active site of this enzyme and write a program which specifies the positions and distances of the most important amino acids as well as important structural features in the vicinity of the active site,” explains acib researcher Christian Gruber.
Once certain characteristics searched are compared with one another, the result is a weighted list of possible candidates that is verified by the researchers based on their enzymological knowledge.
In the next step the most promising candidates are biotechnologically manufactured and tested experimentally in the lab.
The preliminary work on the computer saves on countless experiments and screenings for new enzyme functionality.
For the calculations, the researchers set up a computer cluster with more than 400 cores. This also contains the database that is constantly being expanded.
“Every week about 150 new structures are added,” says Georg Steinkellner from acib, “our system extends itself automatically and searches for new enzyme structures in publicly accessible databases. We also refine the entire system to answer more complex search queries”.
The method was already applied for a patent and published in Nature.