The antibody developability criteria are evolving as well as the format complexity (multispecificity, multivalency, different conformations). Our aim is to be able to predict, design, register, produce and test molecules as fast as possible.

Over the past years, many prediction tools were introduced in our collection. Biophysical properties are predicted by analyzing sequences, computing parameters on structural models, and performing molecular dynamics. However, while progressing in analyzing single binders, the behavior of bi- or tri-specific formats in different flavors remains challenging. Shape, size, and linkers have their importance.

The MOE Antibody Database contains the three-dimensional (3D) structures of antibodies and antigen-antibody complexes derived from the Protein Data Bank (PDB) and provides powerful custom features enabling structural bioinformatics analysis useful for antibody research. We recently used the MOE Antibody Database to identify and analyze the 3D structures of antibodies and complexes that contain disulfide motifs in their CDR-H3s. We present the discovery of disulfide-bonded motifs (CXnC) in the CDR-H3s from the human antibody repertoire sequence data and compared them with structurally known motifs found in antibody structures. These results provide the fundamental understanding on the role of non-canonical cysteines in shaping the complex paratope diversity in human antibodies with potential applications in antibody design and engineering (Prabakaran and Chowdhury, Cell Reports 2020).

Nanobodies, also known as single-domain antibodies, have emerged as a promising class of therapeutics due to their small size, stability, and versatility. However, optimization and design of nanobodies for drug development can be challenging. One of the main aspects is achieving high binding affinity and specificity for the target antigen while maintaining stability and ideal developability. To overcome these challenges, a combination of computational modeling, and experimental validation can be utilized to rationally screen large numbers of candidates and rapidly identify those with the desired properties. During this talk, we will present few use cases of applying general molecular modelling approach and emerging trends in design of nanobody modalities towards the development of nanobody-based therapeutics.

Monoclonal antibodies play an increasingly important role for the development of new drugs across multiple therapy areas. The term ‘developability’ encompasses the feasibility of molecules to successfully progress from discovery to development via evaluation of their physicochemical properties. These properties include the tendency for self-interaction and aggregation, thermal stability, colloidal stability, and optimization of their properties through sequence engineering. Selection of the best antibody molecule based on biological function, efficacy, safety, and developability allows for a streamlined and successful CMC phase. An efficient and practical high-throughput developability workflow (100 s-1,000 s of molecules) implemented during early antibody generation and screening is crucial to select the best lead candidates. This involves careful assessment of critical developability parameters, combined with binding affinity and biological properties evaluation using small amounts of purified material (<1 mg), as well as an efficient data management and database system. Herein, a panel of 152 various human or humanized monoclonal antibodies was analyzed in biophysical property assays. Correlations between assays for different sets of properties were established. We demonstrated in two case studies that physicochemical properties and key assay endpoints correlate with key downstream process parameters. The workflow allows the elimination of antibodies with suboptimal properties and a rank ordering of molecules for further evaluation early in the candidate selection process. This enables any further engineering for problematic sequence attributes without affecting program timelines.

Since the publication of Alphafold demonstrating a “solution” to the protein structure prediction problem the computational biology literature has been awash with novel methods and ideas that promise to revolutionise antibody discovery. In this talk I will discuss our work in areas from accurate rapid antibody structure prediction to binding site prediction and discuss ways to properly assess the wealth of new machine learning methods that are becoming available.

Describing an antibody’s binding site using only one single static structure limits the understanding of the antibody’s function. This limitation is even more pronounced when no experimentally determined structure is available or the crystal structure is distorted by packing effects, which can result in misleading antibody paratope structures. To improve antibody structure prediction and to take the strongly correlated loop and interface movements into account, antibody paratopes should be described as interconverting states in the solution. Therefore, the definition of kinetically and functionally relevant states can be successfully used to improve the accuracy and enhance the understanding of antibody-antigen recognition.

Significant advances in computational methods and hardware accelerated scientific computing have enabled the dawn of a new era of medicine in which lifesaving therapeutic molecules can be designed and optimized with greater speed and precision than ever before. At Johnson & Johnson, we are leveraging data from across the pharmaceutical value chain, manifested in a knowledge graph to inform novel computational, deep learning models to drive innovation and disrupt the therapeutic research and development lifecycle; building and leveraging our collective institutional knowledge across therapeutic programs and indications in order to inform novel AI/ML models to accelerate the development of lifesaving therapies for patients across the globe.

Prescient Design, a Genentech accelerator, is developing integrated methods for optimizing antibody affinity and multiple developability parameters. Key to our efforts are a close integration of antibody engineering, machine learning, and structural biology. Our ML frameworks allow us to integrate later stages of optimization into the earliest stages of discovery, while our high throughput experimental systems allow rapid improvement of all methods and molecules.

This integration starts with the integration of people and scientific culture and ends with tightly integrated computational and experimental systems.

During this talk, I will give an overview of our generative modeling approach, multi-objective optimization, and active learning framework.