Abstract:Predicting the structure of antibody-antigen complexes is a challenging task with significant implications for the design of better antibody therapeutics. However, the levels of success have remained dauntingly low, particularly when high standards for model quality are required, a necessity for efficient antibody design. Artificial intelligence (AI) has significantly impacted the landscape of structure prediction for antibodies, both alone and in complex with their antigens. We utilized AI-guided antibody modeling tools to generate ensembles displaying diversity in the complementarity-determining region (CDR) and integrated those into our previously published AlphaFold2-rescored docking pipeline, a strategy called AI-augmented physics-based docking. We highlight that the quality of the ensemble is crucial for docking performance, that including too many models can be detrimental and that prioritization of models is essential for achieving good performance. In this study, we also compare docking performance with AlphaFold, the new benchmark in the field. We distinguish between two types of success tailored to specific downstream applications: 1) criteria sufficient for epitope mapping, where gross quality is adequate and can complement experimental techniques, and 2) criteria for producing higher-quality models suitable for engineering purposes. Our results robustly demonstrate the advantages of AI-augmented docking over AlphaFold2, further accentuated when higher standards in quality are imposed. Docking performance is noticeably lower than the one of AlphaFold3 in both epitope mapping and antibody design. While we observe a strong dependence on CDR-H3 length for physics-based tools on their ability to successfully predict, this helps define an applicability range where physics-based docking can be competitive to AlphaFold3.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: **How to predict the structure of antibody - antigen complexes more accurately** to support better antibody therapy design. Specifically, the authors focus on increasing the success rate of antibody - antigen complex structure prediction under high - quality standards, which is crucial for efficient antibody design. ### Main problems and challenges 1. **Low prediction quality**: - Current prediction methods have a low success rate under high - quality standards, especially when precise atomic - level structures are required. 2. **Uniqueness of antibody - antigen complexes**: - The structure prediction of antibody - antigen complexes is more complex than that of general protein - protein complexes because of the lack of co - evolutionary constraints, which makes traditional physics - based molecular docking tools (such as AlphaFold) perform poorly in this area. 3. **Flexibility of CDR regions**: - The complementarity - determining regions (CDR) of antibodies, especially the CDR - H3 loop, have a high conformational diversity, which challenges traditional rigid docking methods. 4. **Model selection and priority**: - An excessive number of generated structural models may lead to performance degradation, so how to select and prioritize these models is also a key issue. ### Solutions To solve the above problems, the authors propose a new method that combines artificial intelligence (AI) and physics - based molecular docking, called **AI - Augmented Physics - Based Docking**. The specific steps include: 1. **Generate diverse antibody models**: - Use multiple AI - guided antibody modeling tools (such as IgFold, ABodyBuilder2 and EquiFold) to generate a diverse set of antibody models, especially for the CDR regions. 2. **Integrate into the docking process**: - Integrate these models into the physics - based docking pipeline and re - score them through AlphaFold2 to evaluate the quality of different models. 3. **Optimize model selection**: - Emphasize the importance of the quality of the model set to docking performance and point out that too many models may be harmful. Set a confidence threshold to prioritize high - quality models, thereby improving the overall prediction performance. 4. **Performance evaluation**: - Compare with the latest benchmark tools (such as AlphaFold3) to evaluate the performance of the new method in two application scenarios: one is for epitope mapping, and the other is for antibody design, which requires higher model quality. Through these improvements, the authors demonstrate the advantages of the AI - Augmented Physics - Based Docking method under high - quality standards, especially when the length of CDR - H3 is moderate, its performance can be comparable to that of AlphaFold3.

AI-Augmented Physics-Based Docking for Antibody-Antigen Complex Prediction

Enhanced antibody-antigen structure prediction from molecular docking using AlphaFold2

Towards the accurate modelling of antibody-antigen complexes from sequence using machine learning and information-driven docking

Improving B-cell Epitope Prediction and Its Application to Global Antibody-Antigen Docking.

What has AlphaFold3 learned about antibody and nanobody docking, and what remains unsolved?

ColabDock: Inverting AlphaFold Structure Prediction Model for Protein-Protein Docking with Experimental Restraints

ABAG-docking benchmark: a non-redundant structure benchmark dataset for antibody–antigen computational docking

AI driven approaches in Nanobody Epitope Prediction: Are We There Yet?

Benchmarking AlphaFold‐enabled molecular docking predictions for antibiotic discovery

Benchmarking AlphaFold for protein complex modeling reveals accuracy determinants

Antibody complementarity-determining region design using AlphaFold2 and DDG predictor

Computational Structure Prediction for Antibody-Antigen Complexes From Hydrogen-Deuterium Exchange Mass Spectrometry: Challenges and Outlook

Docking guidance with experimental ligand structural density improves docking pose prediction and virtual screening performance

Accurate prediction of CDR-H3 loop structures of antibodies with deep learning

A Comparison of Antibody-Antigen Complex Sequence-to-Structure Prediction Methods and their Systematic Biases

PoseBusters: AI-based docking methods fail to generate physically valid poses or generalise to novel sequences

Evaluation of AlphaFold2 structures as docking targets

EuDockScore: euclidean graph neural networks for scoring protein-protein interfaces

Antibody-Antigen Docking and Design via Hierarchical Equivariant Refinement

A protocol for information-driven antibody-antigen modelling with the HADDOCK2.4 webserver