DNA and RNA rarely work alone. In living cells, they often team up with proteins to form complexes that help control gene expression, antiviral defence and development. These complexes are central to biology, but they remain hard targets for drug discovery.
Grigory Nikolaev, PhD, from the Laboratory of Bioinformatics and Protein Engineering at the International Institute of Molecular and Cell Biology in Warsaw, will tackle this problem with DeepGeNAPL. His new NCN-funded project will develop an AI tool to study and generate small molecules that target nucleic acid–protein complexes.
The project focuses on DNA–protein and RNA–protein assemblies. These systems matter because they regulate many cellular processes. Yet researchers still lack effective ways to design small molecules that engage them predictably.
“The system will support two linked tasks. First, it will help researchers analyse how small molecules interact with DNA-protein and RNA-protein assemblies. Then it will use that information to guide the design of new molecular candidates. This should make the early search for biologically active compounds more focused, while keeping the claims at the level of discovery support rather than promising new drugs,” says Grigory Nikolaev.
DeepGeNAPL will therefore support an early stage of discovery. It will help researchers explore hard molecular targets, understand possible interactions and select more promising molecular candidates for further study.
The project develops IIMCB expertise at the intersection of artificial intelligence, structural bioinformatics, nucleic acid biology and molecular medicine. It also supports a broader research direction: developing computational tools for biological targets that are important, but hard to study with standard methods.
The project “Deep learning–driven design of small molecules targeting nucleic acid–protein complexes” received 687,092 PLN in funding under the SONATA 21 call.
Funding note
The project described in this article is funded by the National Science Centre, Poland: “Deep learning–driven design of small molecules targeting nucleic acid–protein complexes”, project no. 2025/59/D/ST6/02248, SONATA 21.