Machine Learning for Protein Engineering - Part 1 banner

第2回年次会議「タンパク質工学向け機械学習:パート1」

PEGS Europeの「タンパク質工学向け機械学習」会議では、人工知能と機械学習の応用による、バイオ医薬品の品質・精度・開発可能性の向上を推進する情報科学技術と戦略について2部構成で探求します。2023年の会議は、ユースケースの重要な検討、実験的検証の手順、従来の発見プラットフォームとの統合、データ作成/キュレーションを通じて、発見・最適化のキャンペーンに焦点を当てます。これらのプログラムは、PEGS Europeコミュニティに、この初期段階の技術群の採用におけるこれまでの成功と、研究機関がこの分野への多大な投資の可能性をどのように実現できるかについて議論するフォーラムを提供します。この会議では、これらの技術の重要な価値を示し、理論より実用的な応用に重点を置きます。形式は教育的でインタラクティブなものとなります。

Recommended Short Course*
Monday, 13 November, 14:00 - 17:00
SC1: Machine Learning Tools for Protein Engineering
*Separate registration required. See short courses page for details. All short courses take place in-person only.

11月15日(水)

Registration Open and Morning Coffee07:30

DE NOVO DESIGN USE CASES
de novo設計のユースケース

08:25

Chairperson’s Opening Remarks

Tushar Jain, PhD, Principal Scientist, Computational Biology, Adimab LLC

08:30

Lab-in-the-Loop, an ML-Driven Platform for Automated Molecular Discovery and Design

Nathan Frey, PhD, Machine Learning Scientist, Prescient Design, a Genentech Company

We will discuss the “Lab-in-the-loop” system, a collaboration between Prescient Design and Antibody Engineering at Genentech, to build and integrate state-of-the-art machine learning methods with large molecule design and discovery capabilities. Lab-in-the-loop encompasses generative models, pseudo-oracles, physics-based modeling, large language models, wet-lab assays, and active learning to fundamentally change early-stage drug discovery.

09:00

Generation and Experimental Validation of Novel de novo Abs with Unique Functionalities

Yanay Ofran, PhD, Founder, CEO, Biolojic Design Ltd.

Most therapeutic antibodies are simple antagonists. However, like all proteins, antibodies can be sophisticated nano-machines. Biolojic Design uses AI to program antibodies to become dynamic functional switches affecting biology in new ways. I will describe our AI-design process, and share clinical data from the first AI-designed therapeutic antibody. I will also show preclinical data on multi-specific antibodies illustrating their potential to improve outcome in cancer and autoimmune diseases.

09:30 Talk Title to be Announced

Speaker to be Announced

Session Break to Transition into Plenary Keynote10:00

PLENARY KEYNOTE SESSION
基調講演(プレナリーセッション)

10:10

Introduction

Enkelejda Miho, PhD, Professor, Dean, University of Applied Sciences and Arts Northwestern Switzerland

10:15

Benchmarking the Impact of AI Biologics Discovery and Optimisation for Pharma

Rebecca Croasdale-Wood, PhD, Director, Augmented Biologics Discovery & Design, Biologics Engineering, Oncology, AstraZeneca

The biologics landscape is rapidly changing with the number of AI-enabled biologics in pre-clinical and clinical stages estimated to be 50-60 (1). This change is driven by the increase in enterprise software solutions to capture and store data, augmented discovery workflows, improvements in machine learning technology, and advances in computing power. Augmented biologics discovery has the potential to revolutionize biologics discovery, yet information of how in silico technologies perform, versus traditional discovery platforms is scarce. At PEGS Europe, we will present current in silico biologics design and optimisation technologies, with a focus on our internal efforts to benchmark the impact of combining novel in silico technologies with our existing biologics discovery platforms.

Coffee Break in the Exhibit Hall with Poster Viewing11:00

11:45

Accelerating the Discovery Pipeline with ML: From Library Design to Discovery and Optimization, and Early Developability Screening

Tushar Jain, PhD, Principal Scientist, Computational Biology, Adimab LLC

Incorporating predictive modeling into experimental workflows holds great promise for accelerating discovery, guiding optimization, and prioritizing leads. Here we discuss application of ML to design of synthetic libraries for discovery, hybrid experimental-modeling approaches for selection of functionally diverse antibodies, and developability predictions that decrease resource-intensive experiments. Our integration of ML into a larger informatics/data platform enables predictions to inform candidate selection throughout the discovery and lead optimization process.

12:15

KEYNOTE PRESENTATION: Antibody Structure and Dynamics in Solution

Klaus R. Liedl, PhD, Professor, Head, General, Inorganic, & Theoretical Chemistry, University of Innsbruck

Antibodies are highly flexible molecules, due to the hinge regions, the elbow linkers, the interdomain interfaces between Ig-fold domain pairs and the loops in the paratope. We demonstrated that the binding competent structure is normally the dominant structure in solution, even though it is often not the structure found for unbound antibodies. The resulting opportunities and challenges for AI-driven antibody structure prediction are discussed in the light of these findings.

Sponsored Presentation (Opportunity Available)12:45

Session Break13:15

Luncheon Presentation (Sponsorship Opportunity Available) or Enjoy Lunch on Your Own13:20

Session Break14:20

IMPLEMENTATION CHALLENGES AND SOLUTIONS
実装の課題とソリューション

14:30

Chairperson’s Remarks

Jeffrey Ruffolo, PhD, Machine Learning Scientist, Profluent Bio

14:35

Developing Internal AI Capabilities via External Collaborations and Internal Resources

Hubert Kettenberger, PhD, Head, Computational Protein Engineering, Roche

AI applications have become increasingly powerful, and play an increasing role in today's research and development. At the same time, there is no consensus yet regarding AI methodologies, and how to best integrate them in the discovery and development process. Building internal capabilities and establishing external collaborations can help navigate through this exciting new augmentation of biologics drug development.

15:05

Conformational Entropy as a Potential Liability of Computationally Designed Antibodies

Michele Vendruscolo, PhD, Professor, Biophysics, University of Cambridge

In silico antibody discovery is emerging as a powerful alternative to traditional in vivo and in vitro approaches. Many challenges, however, remain open to enabling the properties of designed antibodies to match those produced by the immune system. I am going to describe the role of conformational entropy, as determined by accurate statistical mechanics calculations, in determining the binding affinity of designed antibodies for their targets.

Sponsored Presentation (Opportunity Available)15:35

Refreshment Break in the Exhibit Hall with Poster Viewing16:05

EMERGING MODELS AND PLATFORMS
新興のモデルとプラットフォーム

17:00

FEATURED PRESENTATION: Generative Antibody Modelling

Charlotte M. Deane, PhD, Professor, Structural Bioinformatics, Statistics, University of Oxford; Chief Scientist, Biologics AI, Exscientia

Here we show that by optimising an inverse folding model specifically for antibody structures, we are able to outperform generic protein models on sequence recovery and structure robustness, with notable improvement on the hypervariable CDR-H3 loop.We also demonstrate the applications of our model to drug-discovery and binder design and evaluate the quality of proposed sequences.

17:30

Integration of Machine Learning, Structural Biology, and Wet Lab Data to Augment Drug Discovery for Autoimmune Diseases

Nathan Higginson-Scott, PhD, CTO, Seismic Therapeutic

We will discuss how Seismic Therapeutic is using its IMPACT platform to integrate machine learning, structural biology, protein engineering and translational Immunology to accelerate the discovery and development of therapeutics for autoimmune diseases, caused by a dysregulated adaptive immune system.

18:00

ML and Computational Tools for Biologic Molecule Design and Insight

Andrew Buchanan, PhD, FRSC, Principal Scientist, Biologics Engineering, Oncology, AstraZeneca

This presentation will overview the need for foundational high quality digital wet-lab and multi-lingual/inter-disciplinary collaboration for antibody and peptide AI ML. It will then discuss FAIRe(nough) data in Research for AI ML applications and a new method for ultra high throughput data generation to enable ML. Finally, it will illustrate progress in benchmarking computational strategies, FEP+ & LLMs, for affinity predictions and successful generative ML models.

18:30

Generative Modeling for Functional Protein Design

Jeffrey Ruffolo, PhD, Machine Learning Scientist, Profluent Bio

Generative language models trained on protein sequences have proven incredibly powerful for protein sequence design. In this talk, we will demonstrate how protein language models enable discovery of diverse proteins, which often function on par with natural counterparts- despite significant deviation in sequence space. Beyond generation of sequences, protein language models are effective zero-shot predictors of fitness, enabling direct optimization of function.

Close of Machine Learning for Protein Engineering - Part 1 Conference19:00


* 不測の事態により、事前の予告なしにプログラムが変更される場合があります。

Choose your language
English



Premier Sponsors

Asimov Inc

Evitria

Immatics-Biotechnologies

Integral-Molecular_NEW

LONZA

Phenomex Logo