Nonspecific binding remains a major developability liability for therapeutic antibodies, with potential consequences for pharmacokinetics and safety. Reliable sequence-based prediction could support earlier decision-making, but remains challenging. Here, we present machine learning models based on protein language model embeddings and biophysical descriptors to predict antibody non-specificity. The models identify the VH domain, heavy-chain CDRs, and isoelectric point as key determinants, achieving up to 71% accuracy.

Protein language models provide a powerful framework for learning sequence–property relationships in antibodies, but their industrial reliability is underexplored. We evaluate state-of-the-art PLMs on internal data from numerous therapeutic programs across three developability assays (PSR, HIC, AC-SINS). Domain-adaptive fine-tuning consistently outperforms pretrained representations, and pretrained sequence likelihoods offer useful unsupervised risk signals. Together, these results show PLMs provide robust, complementary signals for early-stage antibody developability assessment and candidate selection.

AI-driven antibody optimization is often developed in isolation, without experimental benchmarks, so apparent gains obscure real ones. With access to affinity maturation benchmarks for head-to-head evaluation, we investigate where AI can truly optimize existing experimental protocols, across novel target epitopes and distinct antibody backbones and paratopes. Automated curation builds training-ready datasets that continually fine-tune optimization models; variants are measured experimentally, so gains are confirmed and utilized for follow-up iterations.

Developability plays a key role in modern antibody engineering but is costly and time-consuming to characterize in the lab. In this talk, we present a method for constructing datasets that are designed for training new machine learning models to predict features of antibody developability.

• Overview of current tools
• Defining true de novo vs. optimization
• Best tests and metrics of design success
• Open challenges
  

• Performance of commercial vs. open solutions: today and tomorrow
• Business motivations for picking commercial vs open solutions: beyond predictive performance
• In 5 years: towards a database of designed antibodies available off the shelf?
• Will in silico design replace or augment traditional wet-lab discovery? What is next?​

Foundation models trained across protein and molecular data are reshaping how we design therapeutics, not merely predicting properties but optimizing them. I'll show how protein language models power two AI/ML efforts: Raygun, which miniaturizes and redesigns natural proteins to improve developability, and PRECISE, a structure-based platform for drug-target screening and engagement. Together they show how learned representations enable more efficient and effective gene therapy and lead discovery.

Expert-driven epitope selection for in silico protein design is a major bottleneck in DMTA cycles. We introduce Epitope Mapper, a hybrid agentic pipeline that automates this step using focused agents with biological tools. It synthesizes literature, performs biophysical calculations, and proposes ranked epitopes that are consistent with human expertise. The system’s suggestions can be steered by user-provided context, enabling large-scale protein design and accelerating fully automated end-to-end therapeutic discovery.

Achieving developable multispecific antibody and NANOBODY therapeutics requires moving beyond retrospective assessment toward proactive, design-integrated prediction. We describe how machine learning models for key developability properties have been embedded into our multispecific design workflows, enabling earlier and more informed decision-making across therapeutic programs. We discuss large-scale experimental campaigns that underpin these models, sharing practical insights on data strategy, model building, and the unique challenges posed by complex multispecific formats. Project case studies illustrate how computational predictions translate into real design decisions, and we address the critical gaps that currently limit the field.

Antibody developability reflects multiple coupled properties that are difficult to optimize simultaneously. We combine machine learning, antibody repertoire information, and experimental measurements to predict and engineer developability across VHH and IgG formats. Examples include identifying sequence and surface features that drive VHH aggregation, predicting IgG self-association and high-concentration behavior, and using conservative CDR germlining to improve biophysical properties while preserving affinity and function.

We will present an overview of AI-enabled approaches for predicting developability for biologics including multi-specifics. The talk will touch on data foundations, modeling strategies, and workflow integration across biologics discovery, with selected examples showing how in silico methods can support design, chain pairing, lead optimization, and candidate selection while balancing efficacy, developability, and potential immunogenicity considerations.

Generative protein design now reliably produces de novo monospecific antibodies. For many targets, models like JAM-2 generate candidates with drug-quality properties, purely computationally and with high success rates. The next challenge, and where the field is now focused, is extending this success to more complex multifunctional medicines, especially multispecific antibodies. This talk presents examples of zero-shot design and optimization of multispecific antibodies using generative models, and discusses what it will take to bring this same level of reliability to multispecific formats.