A team of students from the Moscow Aviation Institute (MAI), supported by a grant from the Russian Science Foundation, is developing an AI-based software module to assess the risks of taking three or more drugs simultaneously. The work is being conducted at Plekhanov Russian University of Economics. Such combination therapy regimens are common in oncology and in patients with multiple chronic conditions, where the likelihood of side effects increases dramatically.
The module is built on a Bayesian network – a probabilistic model that accounts for multiple factors and their interrelationships. The system analyses drug package inserts from the state medicines registry, evaluating not only direct contraindications but also cases where several drugs affect the same organ, potentially leading to overdose. “After analysis, the system calculates the overall risk,” said the project’s scientific supervisor, Yuri Titov, an associate professor at MAI.
The researchers said the module is straightforward to use. A doctor enters a list of drugs, the patient’s age and sex; the system then checks for drug-drug interactions, assesses the risk of side effects and identifies potentially dangerous combinations. Results are displayed using a traffic-light system: green for compatible, yellow for potential risks, and red for dangerous interactions requiring a change in treatment.
In the first phase of testing, the database was loaded with information on 92 commonly used drugs for chronic heart failure and influenza, and 110 serious side effects that can occur when they are taken together have been identified. The module is currently being tested. In the future, the developers plan to integrate the solution into clinical information systems, including the Unified Medical Information and Analytical System (EMIAS).
According to WHO data, the risk of side effects when taking two to three drugs is about 39%; for four to five drugs, it is 88%; and for six to seven, it reaches 100%. The project’s author, MAI student Nikita Kilmishkin, noted that the module makes the analysis and its results as clear as possible for doctors, helping to reduce the number of adverse drug reactions.
Earlier, researchers at the Kurnakov Institute of General and Inorganic Chemistry of the Russian Academy of Sciences and Lomonosov Moscow State University, with financial support from the Ministry of Education and Science, developed MetalCytoToxDB, the world’s largest database of metal cytotoxicity. It is designed to address a key challenge in developing metal-based anticancer drugs – the lack of systematised data.
