Modern medicines are not only born in pharmacological laboratories, as some imagine. They can be modeled using a supercomputer. Scientists virtually discover and test effective compounds that later become those very same drugs for fighting diseases that were recently considered difficult to cure .

Titans of Computing

Supercomputers perform billions of operations per second. They model reality: they can predict a natural disaster or visualize a car crash test. The decision to use their capabilities to search for new drugs was made after the creation of a supercomputer computing center at the United Institute of Informatics Problems of the National Academy of Sciences of Belarus under the Union State program.

Russia, with its vast experience and capabilities for performing complex technological tasks, remains our direct partner in the field of high technology, says Anna Karpenko, a research fellow at the United Institute of Informatics Problems of the National Academy of Sciences of Belarus:

— Computer modeling helps to weed out “unpromising” compounds and study only effective ones. We don’t need to synthesize or buy a billion of them, we can buy only a hundred and thus save tens of millions of dollars.

At the same time, the time for drug development is significantly reduced, because there is a difference between testing a few compounds or many. Another advantage: a supercomputer can do much more work than a person. The computer modeling method allows us to find a cure for almost any disease: coronavirus, oncology, tuberculosis... Our young scientists have already completed a series of important scientific studies. Their work and calculations are used in projects to create a cure for COVID-19. The resulting compounds are now being tested in China.

Neural networks against cancer

One of the key areas of work of the team of scientists is the search for a universal cure for cancer. The problem is acute. According to WHO, oncology is one of the top 10 causes of death in the world. Belarusian scientists are focusing on the creation of a multi-target drug, Anna Karpenko continues:

— The process begins with identifying a biological target. For example, coronavirus has 28 of them. Each is responsible for different phases of disease development, mutations. If you block their function, you can prevent the disease from progressing. This is especially promising for the fight against oncology, which is known for its large number of mutations. Using a supercomputer and machine learning methods, we are looking for a drug compound that can affect several biological targets at once...