Predicción de proteínas de novo mediante inteligencia artificial para la inhibición del factor NF-κB en el cáncer gástrico

Abstract

In this research, highlights the limited effectiveness of current treatments for gastric cancer and the role of AI tools in the development of new personalized strategies, through the prediction of de novo proteins aimed at inhibiting the Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-κB) associated with this type of cancer. Thirty proteins were predicted, with a similar composition to those stored in standard and experimental databases. Their stability and folding capacity were analyzed based on the energies from intra and intermolecular interactions by molecular dynamics simulations. Furthermore, a molecular docking process was performed between several genes or transcription factors regulated by NF-κB and the predicted proteins, different thermodynamic variables such as Gibbs Free Energy, Dissociation Constant, Enthalpy, Heat Capacity, and Entropy were calculated and compared with respect to the complex of Inhibitor Alpha of kappa B and the protein composed of the p65 and p50 subunits of NF-κB to verify the affinity of their protein-protein interaction and structural conformation state. Finally, the affinity and selectivity of the resulting interactions were evaluated, concluding that the de novo proteins predicted by artificial intelligence are viable for the generation of new treatments and the design of new drugs to treat gastric cancer.