Antigen-Specific Antibody Design via Direct Energy-based Preference Optimization

TL;DR

This paper proposes direct energy-based preference optimization to guide the generation of antibodies with both rational structures and considerable binding affinities to given antigens, and achieves state-of-the-art performance in designing high-quality antibodies with low total energy and high binding affinity simultaneously.

Abstract

Antibody design, a crucial task with significant implications across various disciplines such as therapeutics and biology, presents considerable challenges due to its intricate nature. In this paper, we tackle antigen-specific antibody sequence-structure co-design as an optimization problem towards specific preferences, considering both rationality and functionality. Leveraging a pre-trained conditional diffusion model that jointly models sequences and structures of antibodies with equivariant neural networks, we propose direct energy-based preference optimization to guide the generation of antibodies with both rational structures and considerable binding affinities to given antigens. Our method involves fine-tuning the pre-trained diffusion model using a residue-level decomposed energy preference. Additionally, we employ gradient surgery to address conflicts between various types of energy, such as attraction and repulsion. Experiments on RAbD benchmark show that our approach effectively optimizes the energy of generated antibodies and achieves state-of-the-art performance in designing high-quality antibodies with low total energy and high binding affinity simultaneously, demonstrating the superiority of our approach.

Figures (8)

• Figure 1: The third CDR in the heavy chain, CDR-H3 (colored in yellow), of real antibody (left) and synthetic antibody (right) designed by MEAN MEAN for a given antigen (PDB ID: 4cmh). The rest parts of antibodies except CDR-H3 are colored in blue. The antigens are colored in gray. We use red (resp. black) dotted lines to represent clashes between a CDR-H3 atom and a framework/antigen atom (resp. another CDR-H3 atom). We consider a clash occurs when the overlap of the van der Waals radii of two atoms exceeds 0.6Å.
• Figure 2: Overview of AbDPO. This process can be summarized as: (a) Generate antibodies with the pre-trained diffusion model; (b) Evaluate the multiple types of residue-level energy and construct preference data; (c) Compute the losses for energy-based preference optimization and mitigate the conflicts between losses of multiple types of energy; (d) Update the diffusion model.
• Figure 3: Visualization of reference antibodies in RAbD and antibodies designed by AbDPO given specific antigens (PDB ID: 1iqd (left), 1ic7 (middle), and 2dd8 (right)). The unit of energy annotated is kcal/mol and omitted here for brevity.
• Figure 4: Changes of median CDR $E_{\text{total}}$, $E_{\text{nonRep}}$, $E_{\text{Rep}}$, and CDR-Ag $\Delta G$ (kcal/mol) over-optimization steps, shaded to indicate interquartile range (from 25-th percentile to 75-th percentile).
• Figure 5: A: Tyr (Y) and Phe (F) differ by only one oxygen atom. In contrast, there is a substantial difference between Gly (G) and Trp (W). Gly lacks a side chain, whereas Trp possesses the largest side chain of all amino acids. B: the visualization of the frequency of occurrence of each amino acid at various positions in RAbD CDR-H3 sequences. The sequences are initially aligned using MAFFT katoh2013mafft and subsequently visualized with WebLogo crooks2004weblogo. The width of each column corresponds to the frequency of occurrence at that position.