Publications by authors named "Jose Gutembergue de Mendonca"
Article Synopsis
- Molecular dynamics (MD) simulations generate complex data that are challenging to analyze with traditional methods; however, combining this with deep learning allows for better understanding of structural changes in proteins, such as those caused by mutations.
- The study specifically focuses on the SARS-CoV-2 spike protein's receptor-binding domain (RBD), using distance maps and deep convolutional neural networks to predict how point mutations affect the virus's infectivity and ability to evade immune responses.
- Results showed promising predictive success regarding mutant types that enhance receptor affinity and reduce immunogenicity, with significant correlations found between simulation data and binding free energy changes, potentially aiding in identifying more infectious and immune-evasive strains of the virus.
