mRNA-based vaccines and therapeutics are increasingly popular and used for a variety of conditions. A key challenge in designing these mRNAs is sequence optimization. Even small proteins or peptides can be encoded by a vast number of mRNA sequences, each affecting properties such as expression, stability, and immunogenicity. To facilitate the selection of optimal sequences, we developed CodonBERT, a large language model (LLM) specifically for mRNAs.

AlphaFold2 has transformed structural biology with groundbreaking advances in protein structure prediction. However, despite these advances, many challenges remain. In this talk, I will present a progress update on AlphaFold2 and share insights gained from OpenFold, an optimized and trainable variant of AlphaFold2. I’ll also explore potential paths to address the limitations of AlphaFold-like systems.

Developing effective monoclonal antibody (mAb) therapies requires optimizing multiple properties, known as 'developability,' to ensure they can progress through the development pipeline. However, there is limited understanding of how the characteristics (redundancy, predictability, sensitivity) of developability parameters (DPs) in human-engineered antibodies compare to those in natural antibodies. We analyzed 86 DPs across two million antibody sequences, finding key differences in the predictability and sensitivity of sequence- and structure-based DPs. Our findings reveal that human-engineered antibodies occupy a narrower space within the natural antibody landscape, offering a foundation for more precise mAb design.

Chemical degradation poses significant risks in the developability of antibodies, potentially leading to loss of binding or liabilities like aggregation and immunogenicity. Here, I will present a machine-learning-driven approach to identify structural features crucial for chemical stability. I will share experimental evidence showing that targeted point mutations can effectively mitigate chemical liabilities without compromising binding, and discuss how these insights can inform the multi-parameter optimization of antibodies for enhanced stability.

This talk highlights the integration of strategic data collection and intelligent experimental design to advance AI-powered antibody discovery and optimization. We will share updates from the AIntibody competition, a benchmarking initiative engaging the biotech, pharma, academia, and AI communities. These efforts, alongside tailored discovery campaigns for individual collaborators, aim to accelerate innovation and drive progress in early-stage therapeutic discovery.

Structure-based vaccine design campaigns that aim to stabilize full-length proteins will not succeed when the virus is evading immune response by sequence diversity. In this work, we capitalized on the recent major advanced that have been achieved in protein design using generative AI tools. We in silico designed de novo proteins that scaffold sequence-conserved epitope regions of the antigens of interest. Next, we incorporated the in silico designed mini proteins onto self-assembling nanoparticles and performed pre-clinical animal immunization studies eliciting immune responses.

The adaptive immune response relies on B-cell receptors (BCRs) for pathogen neutralization, yet designing BCRs de novo remains challenging due to structural complexity. Here, we introduce Antibody Generative Pretrained Transformer (AbGPT), a fine-tuned model from a foundational protein language model. Using a tailored generation and filtering pipeline, AbGPT generated 15,000 high-quality BCR sequences, effectively capturing the intrinsic variability and conserved regions critical to antibody design.

Language models learn powerful representations of protein biology. We introduce a new foundation model suite that directly investigates scaling effects for protein generation. We then apply this for applications in antibody and gene editor design.

Many researchers are trying to replicate the success of machine learning (ML) in computer vision and natural language processing in modeling biophysical systems. As a discipline, ML heavily relies on low-cost empirical benchmarks to guide algorithm development, but available benchmarks for biophysical applications have major shortcomings. Drawing inspiration from mutational landscape models, we propose Ehrlich functions, a new class of test functions for biophysical sequence optimization algorithms.

This presentation will explore the development of machine learning models for protein property prediction. Focusing on customized protein language models for the NANOBODY modality, it will demonstrate the implementation of a downstream thermostability predictor. Using this example, the talk will emphasize the critical importance of meaningful data splits in model training and validation.

Adaptyv accelerates data generation for training and validating AI models with a high-throughput lab that companies can access via our software interface and API. We empower protein-design teams to validate their AI models many times faster than before, without the need to run in-house wet labs. We're partnering with dozens of companies-from techbio startups to major pharma-and generated lab data for thousands of novel proteins.

At IPI, we perform 300 antibody and VHH discovery campaigns per year against human targets. Our end-to-end antibody discovery platform integrates antigen production, yeast display libraries (Fab and VHH), next-generation sequencing, antibody production, and biophysical characterization. With a success rate exceeding 85%, this platform generates antibodies that benefit the community. Also, the data sets offer powerful opportunities for AI and machine learning applications, driving innovation in antibody discovery and development.

The IMMREP23 competition evaluated TCR-pMHC interaction prediction methods from 53 participating teams submitting 398 sets of predictions. Results showed reasonable performance for "seen" pMHC targets but near-random performance for "unseen" peptides, highlighting an unsolved generalization challenge. Here we discuss what has been learned from detailed analysis of predictions and provide insights for improving future benchmarks by carefully addressing biases in dataset construction.

The DTU Arena for Life Science Automation (DALSA) is a pioneering initiative aimed at advancing the technologies and use of automation within Life Science R&D and manufacturing for academia as well as industry. Our mission is to establish a central and open access hub for technology development and educational advancement in life science automation, serving as a test bed for the full-scale implementation of DALSA. With successful multi-disciplinary collaborations between academia and industry and involving electrical, mechanical, bio, and computational engineers, we are already seeing the benefits of a centralised open access automation hub.