Machine Learning Approaches for Protein Engineering: Part 2 banner

The use of generative AI, deep learning and machine learning tools are poised to make a tremendous impact on the biologics pipeline, but there need to be guidelines put in place to ensure that the results are measurably improved in terms of investment of time, money, and output. Part 2 of the Machine Learning track at PEGS Europe examines and measures the real impact of using these techniques by implementing benchmarks, experimental validation, standards, and controls to guide the process, and case studies comparing traditional approaches with newer, machine learning approaches will shed light on how to adapt and utilize them for biotherapeutic discovery, prediction, developability, simulation and optimization.

Scientific Advisory Board:

M. Frank Erasmus, PhD, Head, Bioinformatics, Specifica, Inc.
Victor Greiff, PhD, Associate Professor, Immunology, University of Oslo
Maria Wendt, PhD, Global Head and Vice President, Digital and Biologics Strategy and Innovation, Sanofi

Recommended Short Course*
Monday, 4 November, 14:00 – 17:00
SC4: In silico and Machine Learning Tools for Antibody Design and Developability Predictions
*Separate registration required. See short courses page for details. All short courses take place in-person only.

Thursday, 7 November

07:30Registration and Morning Coffee

ADVANCED AI TECHNIQUES FOR ANTIBODY ENGINEERING & DEVELOPMENT

08:55

Chairperson's Remarks

M. Frank Erasmus, PhD, Head, Bioinformatics, Specifica, an IQVIA business

09:00

Artificial Intelligence Supports Antibody Discovery in Dengue

Enkelejda Miho, PhD, Professor, University of Applied Sciences and Arts Northwestern Switzerland; Managing Director, aiNET

Dengue virus poses a serious threat to global health and there is no specific therapeutic for it. The antibody response in dengue infection and immunisation can be deconvoluted with high-throughput sequencing and artificial intelligence methods. Machine learning applied to sequencing data identifies rare and underrepresented dengue-specific antibodies.

09:30

Using ML to Enable Patient-Led Antibody Discovery

Laura S. Mitchell, PhD, Principal Bioinformatician, Alchemab Therapeutics

Foundation models have had a transformative impact across many fields, through their ability to learn from large unstructured datasets, and to be fine-tuned for specific tasks. Here I will introduce the Alchemab discovery platform, which blends computational and experimental approaches at every step. Three foundation models trained on antibody sequences (AntiBERTa, AntiBERTa2-CSSP and FabCon) enable many steps in our discovery process. Our AntiBERTa models achieve state-of-the art predictions in amino-acid-level and sequence-level tasks, while FabCon can generate human-like and developable antibody sequences.

10:00Coffee Break in the Exhibit Hall with Poster Viewing

INNOVATIONS IN HIGH-THROUGHPUT SCREENING, OPTIMISATION, AND ML-DRIVEN SUCCESS PREDICTIONS

10:44

Chairperson's Remarks

M. Frank Erasmus, PhD, Head, Bioinformatics, Specifica, an IQVIA business

10:45

Multi-Modal Learning of Protein Properties

Tunca Dogan, PhD, Professor, Department of Computer Science and AI Engineering, Hacettepe University, Turkey

The identification of the specific functions of each protein is essential for understanding the underlying mechanisms of life and developing novel treatments against deadly diseases. Large language models (LLMs) have emerged as a reliable tool for uncovering hidden knowledge in sequence-based data. In this seminar, I’ll present our work on protein foundation models, which employ LLMs and other deep-learning architectures to embed proteins in high-dimensional vector spaces and learn their functional properties proficiently. To accomplish this, we utilized multi-modal learning, wherein the amino acid sequence, molecular interaction, and text-based data are integrated to construct a holistic representation learning model.

11:15

Machine Learning–Driven Design and Optimisation of Antibodies

Lin Li, PhD, Senior Staff Member, Lincoln Laboratory, Massachusetts Institute of Technology

The design and discovery of early-stage antibody therapeutics is time- and cost-intensive. I will present an end-to-end machine learning–driven single-chain variable fragments (scFv) design framework that uniquely combines large language models, Bayesian optimisation, and high-throughput experimentation. The method enables rapid and cost-effective design of thousands of scFvs across all complementary determining regions. The designed antibodies exhibit strong binding affinities, at high levels of diversity, to a given antigen.

11:45

What Really Happens During a Discovery Campaign? And Can AI/ML Help?

M. Frank Erasmus, PhD, Head, Bioinformatics, Specifica, an IQVIA business

Robust datasets are essential for efficient training of machine learning algorithms, particularly in the context of affinity and epitope prediction. We have developed an iterative selection strategy for yeast equilibrium sorting paired with NGS that promotes recovery of antibody sequences with broad ranges of paratopes and affinities. Coupling these outputs with high-throughput functional screening assays has the potential to yield broadly distributed, validated sequences, ideal for model training.

12:15 Poster Highlight:

IgBlend: Unifying 3D Structure and Sequence for Antibody LLMs

Cedric Malherbe, PhD, Senior AI Scientist, AstraZeneca

Large language models (LLMs) trained on antibody sequences have shown significant potential in the rapidly advancing field of machine learning-assisted antibody engineering and drug discovery. However, current antibody LLMs often overlook structural information, which could enable the model to more effectively learn the functional properties of antibodies. In response, we introduce IgBlend, a LLM which integrates both the 3D coordinates of the backbone and antibody sequences

12:45Luncheon in the Exhibit Hall with Last Chance for Poster Viewing

13:55

Chairperson's Remarks

Maria Wendt, PhD, Global Head and Vice President, Digital and Biologics Strategy and Innovation, Sanofi

14:00

Going for Gold – Update on Sanofi’s Biologics AI Moonshot

Maria Wendt, PhD, Global Head and Vice President, Digital and Biologics Strategy and Innovation, Sanofi

CUTTING-EDGE DEVELOPMENTS IN DE NOVO DESIGN FROM SEQUENCE & STRUCTURE

14:30

Modular Binding Proteins: Combining Machine Learning, Structural Biology, and Experimental Evolution

Andreas G. Plueckthun, PhD, Professor and Head, Biochemistry, University of Zurich

We challenge the paradigm of selection from large universal libraries to obtain binding proteins rapidly and efficiently. For linear epitopes, we found it to be possible to exploit the periodicity of peptide bonds and create a completely modular system, based on a binding protein design that shares the same periodicity. To reach selective and sequence-specific binding, we found it to be advantageous to combine machine learning, structural biology, and experimental evolution.

INTERACTIVE DISCUSSIONS

15:00Interactive Discussions

Interactive Discussions are informal, moderated discussions, allowing participants to exchange ideas and experiences and develop future collaborations around a focused topic. Each discussion will be led by a facilitator who keeps the discussion on track and the group engaged. To get the most out of this format, please come prepared to share examples from your work, be a part of a collective, problem-solving session, and participate in active idea sharing. 

TABLE 4:

Delivering on the AI Antibody Promise: the AIntibody Benchmarking Competition

Andrew R.M. Bradbury, MD, PhD, CSO, Specifica, an IQVIA business

M. Frank Erasmus, PhD, Head, Bioinformatics, Specifica, an IQVIA business

  • AI promises in antibody discovery and optimization: will they really revolutionize the field? Or just another way of addressing solved problems?
  • What can AI do now? And where are we seeing the greatest value relative to existing technologies? 
TABLE 3:

Machine Learning for MHC Peptide Presentation and Antibody Immunogenicity Prediction

Mojtaba Haghighatlari, PhD, Senior Machine Learning Scientist, Pfizer Inc.

  • Novel deep learning approaches for predicting MHC antigen presentation and the modeling challenges
  • Interpretability and explainability of the available deep learning models
  • Best practices in data preparation for machine learning of peptidomics datasets
  • Antibody design by transitioning from peptide presentation to protein screening

15:40Close of PEGS Europe Summit