Computational design of developable therapeutic antibodies: efficient traversal of binder landscapes and rescue of escape mutations

Frédéric A. Dreyer,Constantin Schneider,Aleksandr Kovaltsuk,Daniel Cutting,Matthew J. Byrne,Daniel A. Nissley,Newton Wahome,Henry Kenlay,Claire Marks,David Errington,Richard J. Gildea,David Damerell,Pedro Tizei,Wilawan Bunjobpol,John F. Darby,Ieva Drulyte,Daniel L. Hurdiss,Sachin Surade,Douglas E. V. Pires,Charlotte M. Deane

DOI: https://doi.org/10.1101/2024.10.03.616038

2024-10-04

Abstract:Developing therapeutic antibodies is a challenging endeavour, often requiring large-scale screening to produce initial binders, that still often require optimisation for developability. We present a computational pipeline for the discovery and design of therapeutic antibody candidates, which incorporates physics- and AI-based methods for the generation, assessment, and validation of developable candidate antibodies against diverse epitopes, via efficient few-shot experimental screens. We demonstrate that these orthogonal methods can lead to promising designs. We evaluated our approach by experimentally testing a small number of candidates against multiple SARS-CoV-2 variants in three different tasks: (i) traversing sequence landscapes of binders, we identify highly sequence dissimilar antibodies that retain binding to the Wuhan strain, (ii) rescuing binding from escape mutations, we show up to 54% of designs gain binding affinity to a new subvariant and (iii) improving developability characteristics of antibodies while retaining binding properties. These results together demonstrate an end-to-end antibody design pipeline with applicability across a wide range of antibody design tasks. We experimentally characterised binding against different antigen targets, developability profiles, and cryo-EM structures of designed antibodies. Our work demonstrates how combined AI and physics computational methods improve productivity and viability of antibody designs.

Bioinformatics

What problem does this paper attempt to address?

The paper attempts to address several key challenges in therapeutic antibody design. Specifically, the research team developed a computational pipeline aimed at efficiently generating therapeutic antibody candidates with high affinity and developability by combining physical methods and artificial intelligence techniques. The study mainly focuses on the following aspects:

1. **Efficiently traversing the binding epitope landscape**: The research team demonstrated how to identify antibodies that are highly different in sequence while maintaining binding capability to the original strain (e.g., Wuhan strain).

2. **Recovering binding capability from escape mutations**: For escape mutations on the antigen, the research team showed how to redesign antibodies to regain binding affinity to new variants.

3. **Improving antibody developability characteristics**: The research team proposed a method to improve the developability characteristics of antibodies, such as reducing aggregation propensity and increasing thermal stability, while retaining binding properties.

The effectiveness of a small number of designed antibodies against multiple SARS-CoV-2 variants was experimentally validated, demonstrating their performance in different tasks, including traversing the binding epitope landscape, recovering binding capability, and improving antibody developability. These results collectively showcase an end-to-end antibody design pipeline suitable for a wide range of antibody design tasks, with experimental characterization of binding, developability profiles, and cryo-EM structures of designed antibodies against different antigen targets.