Graphen Artificial Intelligence
Optimize vaccines based on variant information, structure energy and contact information.
Developing an efficacious vaccine during a pandemic is critical in providing the global community with immune protection.
Atom used a bioinformatic approach with deep learning methodology to aid in designing an epitope peptide-based vaccine against the protein of the virus. We combined the knowledge of epidemiology and structural biology. On one hand, we monitor the variants of virus to find the spectrum immunity regions; on the other hand, we use our energy model to find the most suitable for BCR or TCR interaction on the antigen. Additionally, we take into consideration of the neutralized antibody-antigen complex to optimize the epitope of representation of vaccine structure. Lastly, simulating the structure of vaccine candidate to verify the structure similarity with our protein structure simulation module.
