Antigen-Specific Antibody Design and Optimization with Diffusion-Based Generative Models for Protein Structures

By Shitong Luo, Yufeng Su, Xingang Peng, Sheng Wang, Jian Peng, Jianzhu Ma. bioRxiv, 2022

305 citations recorded by the source corpus.

Antibodies are immune system proteins that protect the host by binding to specific antigens such as viruses and bacteria. The binding between antibodies and antigens is mainly determined by the complementarity-determining regions (CDR) of the antibodies. In this work, we develop a deep generative model that jointly models sequences and structures of CDRs based on diffusion probabilistic models and equivariant neural networks. Our method is the first deep learning-based method that generates antibodies explicitly targeting specific antigen structures and is one of the earliest diffusion probabilistic models for protein structures. The model is a “Swiss Army Knife” capable of sequence-structure co-design, sequence design for given backbone structures, and antibody optimization. We conduct extensive experiments to evaluate the quality of both sequences and structures of designed antibodies. We find that our model could yield competitive results in binding affinity measured by biophysical energy functions and other protein design metrics.

Subfields: Biotechnology, Molecular Biology, Radiology, Nuclear Medicine and Imaging

View Paper: https://www.semanticscholar.org/paper/37355fe82b7a9cf96ead194018b1775eec9af605

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

In Radiology, Nuclear Medicine and Imaging

Citations in the corpus

SE(3) diffusion model with application to protein backbone generation (2023, methodology): “Other recent works have attempted a generative modeling on rotations through an iterative denoting paradigm akin to diffusion modeling in applications to protein modeling (Anand & Achim, 2022; Luo et al., 2022), as well as robotics (Urain et al., 2022).”
Diffusion probabilistic modeling of protein backbones in 3D for the motif-scaffolding problem (2022, unknown): “Previous works have addressed this challenge through brute force by sampling multiple lengths and placements, and relied on post-hoc filtering to identify the most promising scaffolds (Wang et al., 2022; Yang et al., 2021).”
InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation (2023, background): “…Models Diffusion models [16, 17, 25, 26, 27, 28, 29, 30, 31] have achieved unprecedented results in various generative modeling tasks, including image/video generation [9, 32, 33, 34, 35], audio generation [36], point cloud generation [37, 38, 39, 40], biological generation [34, 41, 42, 43], etc..”
Broadly applicable and accurate protein design by integrating structure prediction networks and diffusion generative models (2022, unknown): “Recent work has sought to adapt DDPMs for protein monomer design by conditioning on small protein “motifs” 2,6 or on secondary structure and adjacency (“fold”) information 5 .”
Diffusion Models: A Comprehensive Survey of Methods and Applications (2022, background): “Experiments show that DiffAb can be used for various antibody design tasks, such Manuscript submitted to ACM Yang et al. as jointly generating sequence-structure, designing CDRs with fixed frameworks, and optimizing antibodies.”
DiGress: Discrete Denoising diffusion for graph generation (2022, background): “…works have recently proposed discrete diffusion models for text, images, audio, and attributed point clouds (Austin et al., 2021; Yang et al., 2022; Luo et al., 2022 Apart from diffusion models, there has recently been a lot of interest in non-autoregressive graph generation using VAEs, GANs or…”
Structure-based drug design with equivariant diffusion models (2022, unknown): “This approach is inspired by (Luo et al., 2022) but note this does not allow for direct optimization of specific properties.”