At AItenea Biotech, in collaboration with several CSIC research centers and Comillas Pontifical University, we’ve contributed to the development of an innovative AI-powered molecular design protocol. This technology has enabled us to create a new molecule from scratch, showing promising properties for the treatment of diseases such as Alzheimer’s and certain types of cancer.
Targeting a Key Protein
The focus of the research is DYRK1A, a protein linked to conditions like Alzheimer’s, Down syndrome, and specific forms of cancer. Inhibiting this protein could help slow down harmful processes like the buildup of hyperphosphorylated tau proteins, which are closely associated with neurodegeneration.
By combining traditional techniques with advanced AI models, we successfully designed a molecule that strongly inhibits DYRK1A. It also shows antioxidant and anti-inflammatory properties and, crucially, it can cross the blood-brain barrier—an essential feature for any central nervous system therapy.
AI Transforms Drug Discovery
The breakthrough comes from a generative computational model capable of designing thousands of potential drug candidates—even when experimental data is limited. The model is based on graph-based molecular representations and structural motifs, allowing for more accurate predictions of biological activity.
We also integrated AI-enhanced QSAR models to predict the efficacy and potential toxicity of each compound. This hybrid approach significantly accelerates the drug discovery process compared to conventional methods.
Towards a New Generation of Therapies
The most promising compounds have already been synthesized and tested in enzymatic and cellular assays. The next phase will involve optimizing their properties and evaluating them in preclinical models.
At the same time, we are refining the protocol by merging generative and predictive models and exploring new techniques such as reinforcement learning to further enhance the molecule design process.
This achievement not only opens new possibilities for treating neurodegenerative and oncological diseases, but also lays the foundation for applying this protocol to a wide range of therapeutic targets. It’s a step forward toward faster, more precise, and more personalized medicine—powered by artificial intelligence.
🔬 Want to dive deeper into the research?
Almudena García
