Expert talks about the effect of AI in the drug development process

The introduction of artificial intelligence (AI) into the drug development process allows to speed up processes and improve the accuracy of research, Artur Kadurin, head of the Deep Learning in Life Sciences group at AIRI Institute, told Izvestia on February 12.

"First of all, this applies to preclinical research. AI-based tools allow analyzing huge amounts of data, including genomic, proteomic and metabolomic data. Thanks to this, researchers can identify potential targets for synthesizing new drugs faster and spend fewer resources on experiments," he specified.

According to him, machine learning is also used to predict the interaction of molecules with biological targets, reducing the time to find promising compounds that could become the basis for drugs.

"The molecules that could become therapeutic in theory are much more than hundreds of billions, and instead of synthesizing each one in the laboratory, 'digital' experiments can be conducted at the first stage," Kadurin added.

The expert explained that in the clinical trials phase, AI-based tools are useful for analyzing results, looking for patterns in the data. In addition, data analysis tools can be useful for finding new applications for existing drugs.

"One of the major trends in current research on the application of AI in pharmacology is the design of drug molecules with defined properties. This process is partly similar to the search for new materials: by designing molecules in simulations, scientists find variants that are most likely to behave in a certain way to become the basis for new drugs in the future," Kadurin said.

Speaking about the most important international developments for this field, he noted DeepMind's AlphaFold, which has revolutionized the understanding of protein structure. The tool predicts the three-dimensional structure of proteins with high accuracy, as many drugs target proteins specifically.

"In Russia, there are many scientific groups engaged in chemoinformatics research, including the Center for Computational Physics at MIPT, the Deep Learning in Life Sciences research group at AIRI, research teams at the National Research University Higher School of Economics, ITMO, Innopolis University and other organizations. In addition, the Skolkovo Innovation Center supports several projects related to the application of AI in pharmacology. Among them are developments in the field of big data analysis and the search for new drug targets," the expert added.

According to him, the study of AI application in pharmacology allows reducing the number and cost of laboratory tests in the process of developing new drugs.

"Despite this, AI alone cannot create a market-ready drug. AI-based tools are among the many programs used by scientists that speed up the process, but human oversight remains indispensable at all stages of drug development: conducting experiments in the physical world, validating data, interpreting results, conducting clinical trials and complying with regulatory requirements are completely controlled by humans," Kadurin concluded.

Earlier, on December 26, 2024, Russian Deputy Prime Minister Dmitry Chernyshenko said that Russia is one of the world leaders in the development of artificial intelligence technologies. According to him, Russian scientists have unique neural processors and mathematical models.