AbODE: Ab Initio Antibody Design using Conjoined ODEs
This work addresses the problem of accelerating vaccine discovery by generating new antibodies that target specific antigens, representing a domain-specific advancement.
The paper tackled the problem of de novo antibody design by co-designing amino acid sequences and 3D structures, overcoming challenges from protein folding, inverse folding, and docking. The result was that the proposed model AbODE significantly outperformed existing methods on standard metrics across benchmarks.
Antibodies are Y-shaped proteins that neutralize pathogens and constitute the core of our adaptive immune system. De novo generation of new antibodies that target specific antigens holds the key to accelerating vaccine discovery. However, this co-design of the amino acid sequence and the 3D structure subsumes and accentuates some central challenges from multiple tasks, including protein folding (sequence to structure), inverse folding (structure to sequence), and docking (binding). We strive to surmount these challenges with a new generative model AbODE that extends graph PDEs to accommodate both contextual information and external interactions. Unlike existing approaches, AbODE uses a single round of full-shot decoding and elicits continuous differential attention that encapsulates and evolves with latent interactions within the antibody as well as those involving the antigen. We unravel fundamental connections between AbODE and temporal networks as well as graph-matching networks. The proposed model significantly outperforms existing methods on standard metrics across benchmarks.