Instead of Thousands of Test Tubes: Skoltech Startup Uses AI to Speed Up Drug Discovery
Russian scientists have launched a platform that combines AI and molecular modeling for drug development. It helps researchers design molecules, test how they interact with proteins, and identify the most promising candidates. The platform is currently undergoing testing.
Developing a new drug is a long, expensive, and risky process. On average, it takes 10 to 15 years for a drug to go from an initial idea to the market, while developing a single molecule can cost $1 billion to $2 billion. Every month, and sometimes every week, of delay means thousands of people do not receive treatment in time. That is why tools that can accelerate the early stages of drug development are particularly valuable.
Ligand Pro, a Russian startup founded by researchers from Skoltech, has introduced a platform for developing drug molecules using artificial intelligence and molecular modeling. The software was presented in late September at the Moscow Startup Summit.
Thousands of Hours of Manual Work Are No Longer Needed
The Ligand Pro platform brings together in a single digital environment tools that researchers previously had to use separately. The system helps them formulate and test molecular design hypotheses, generate new structures, and optimize existing compounds.
The platform can also predict how a molecule may interact with a target protein and assess its physicochemical properties and structural risks. Dedicated tools can analyze the patent landscape and evaluate the synthetic accessibility of compounds – in other words, how feasible it is to produce a particular molecule in the laboratory.
As a result, researchers no longer have to work through every option manually. Instead, they can create a shortlist of the most promising molecules and send them on for synthesis and biological testing.
A Unified Working Environment
Another advantage of the startup's platform is its versatility. Drug development teams today work with around a dozen different software tools. One calculates how a molecule binds to a protein. Another predicts its properties. A third generates new molecular structures. The central problem is that these programs are not connected to one another.
Researchers have to transfer data manually from one program to another, compare results, and keep track of a large amount of contextual information – for example, why they chose one molecule rather than another. That introduces the potential for human error and consumes considerable time on routine work.
The Skoltech team set out to address this problem. Sergey Nikolenko, who leads the project's cheminformatics work, explains: “We wanted to create not a catalog of online tools, but a unified molecular design environment. A researcher can formulate a hypothesis, associate a series of molecules with it, test the hypothesis through computational analysis, and then select candidates based on the combined results.”
Saving Time and Resources
The main benefit Ligand Pro can offer is savings in both time and money. The startup's platform allows researchers to eliminate clearly unsuccessful options at the computational stage and focus their attention on promising candidates worth pursuing.
Pharmaceutical companies could significantly reduce spending on preclinical research, while research teams could test their hypotheses and publish their results more quickly. It is important, however, to understand exactly where the platform can save time: it accelerates the early stages of drug development, specifically the discovery and optimization of molecules. It cannot replace or accelerate clinical trials.
A First for Russia
Notably, Russia has not previously had a platform of this kind. Ligand Pro is the country's first example of such software. Russian researchers previously had to either use foreign platforms or piece together a workflow from separate tools. They now have a domestic environment that can take them through the entire molecular design process – from an initial hypothesis to a shortlist of candidates. For the pharmaceutical industry, this represents a step toward reducing dependence on external services. For Russia's IT sector, it is a distinctive example of a product that brings together AI, biology, and chemistry.
Potential for Export
Drug development is one of today's most global challenges. Pharmaceutical companies around the world are looking for ways to make its early stages faster and less expensive. If the Russian platform demonstrates its effectiveness, it is expected to find demand in countries developing their own pharmaceutical industries, including China, India, Brazil, and Pakistan.
The Skoltech team is currently gathering feedback on the startup's platform from pharmaceutical companies and research groups, while adding new computational modules. Once testing is complete, the team plans to open the system to a broader range of users.
