ITMO: DiZyme Platform for predicting the catalytic activity of artificial enzymes

Main article: Nanoparticles in Medicine

2022: Platform Presentation DiZyme

On January 21, 2022 University ITMO , he announced that his scientists had created a platform for predicting catalytic activity - nanozim artificial enzymes. According to the information company, in a couple of seconds algorithm it determines all the main parameters of the reaction and prompts the optimal conditions for its implementation. In the future, the resource will help in the development of new drugs and diagnostic systems.

Quantitative prediction of catalytic activity of nanozyme

Unlike natural enzymes, nanozymes are much more stable (not requiring storage conditions), much cheaper, which reduces production costs, and easy to produce. Therefore, they are widely used in various fields: from cancer diagnosis and treatment to environmental protection and biosensory. If you learn to predict which nanozyms will achieve high results, you can significantly accelerate the process of obtaining new materials and biochemical systems. However, until now there was no tool that would accurately predict the course of enzymatic reactions on nanoparticles.

Scientists at ITMO University solved the problem by creating an open extensible platform. DiZyme It contains a database of nanoparticles with enzymatic activity, an interactive data visualizer and artificial intelligence algorithms that predict the parameters of the enzymatic reaction, that is, its speed and affinity of substances involved in the chemical process.

"Our service focuses primarily on investigating nanoparticles with peroxidase activity - a subclass of enzymes that catalyze the oxidation of substrates by hydrogen peroxide. On these systems, we have a large array of data. Anyone can go to the site, enter the chemical formula and parameters of the nanoparticle of interest, after which the algorithm predicts its activity. The whole process takes a few seconds, "- emphasizes the first author of the study, ITMO master's student Yulia Razlivina.

The service will be useful to different users. As those who want to develop a specific material for pressing problems: for this, it is enough to inscribe a chemical formula, after which the algorithm automatically evaporates all other parameters and gives out the boundaries of the activity of nanoparticles of this composition. So do those who are seriously involved in this subject area.

"Moreover, the service is not limited to peroxidase activity alone and could potentially be transferred to all other types, which is what is planned to do when enough experimental data is digitized. Our task is to significantly accelerate experimental research and minimize the costs of it. The more scientists use our platform, the better it gets, " adds one of the study authors, ITMO University graduate student Nikita Serov.

Anyone can enter the parameters of their system to obtain graphs of the activity of nanoparticles from various indicators, for example, pH or temperature. Thus, scientists will be able to predict the enzymatic activity of the nanomaterial before conducting experiments. In normal practice, this is measured manually and requires a large amount of time and verification empirically.

"It was important for us to demonstrate the accuracy of the model. To do this, we selected 16 different samples obtained in our laboratory that measured peroxidase activity to compare how accurate platform predictions are. It turned out that about 70% of the systems, the service showed a very good match in terms of parameters. For the remaining 30%, values ​ ​ fall within the permissible limit of accuracy for evaluation before conducting experiments. This suggests that our platform has a high level of accuracy in predicting enzymatic activity, " commentates Doctor of Art, Head of the SCAMT Institute of ITMO University Vladimir Vinogradov.

Links

Platform Portal DiZyme