ITMO University has created an AI-based algorithm, which significantly accelerates the search for promising compounds for autoimmune drugs development. The technology reduces the multi-year process of manually selecting molecules to several days, the university’s press service told TASS. The researchers plan to hold experiments to test the efficacy and safety of the selected molecules in vitro and in vivo. According to them, the new method can be used to search for inhibitors of other therapeutic targets.
Scientists at the ITMO Center for Artificial Intelligence in Chemistry used machine learning methods to search for spleen tyrosine kinase inhibitor (Syk), a protein that plays a key role in the development of immune thrombocytopenia. This is an autoimmune disorder characterized by a low platelet count, which causes spontaneous bleeding.
The treatment is based on spleen tyrosine kinase inhibitors. Special compounds suppress the activity of the protein, eliminating the symptoms of the disease. Not all existing Syk inhibitors are effective. Fostamatinib, the approved drug for the treatment of immune thrombocytopenia, may cause side effects.
Scientists are looking for new suitable Syk inhibitors. However, the traditional search for such compounds requires manual testing of thousands of molecules and can take years. The new method offered by ITMO specialists solves the problem of finding spleen tyrosine kinase inhibitors by generating molecules with specified properties and quickly selecting the most promising options.
The researchers have used the new method to identify 139 candidate molecules that inhibit the Syk protein, which seem more promising than the existing drugs.
Earlier it became known that Russia has developed and successfully tested a quantum algorithm of generative artificial intelligence for use in chemistry and drug development.
