Conference Day One | Wednesday, October 15

For more information, download the event guide.

8:00 am Check-In & Light Breakfast

8:50 am Chair’s Opening Remarks

Strategies to Obtain High Quality Datasets to Enhance Validation of Models

9:00 am Building Purposeful Datasets to Reduce Model Bias & Improve Predictive Accuracy

  • Ryan Emerson Senior Vice President & General Manager - Data Science & Platform Solutions, A-Alpha Bio Inc.

Synopsis

  • Learn why retrofitting existing experimental data for model training often introduces intrinsic bias that skews outcomes
  • Explore strategies for designing data collection protocols explicitly for machine learning, ensuring consistency and reducing noise
  • Understand the importance of resource investment in systematic data generation and where to prioritize for maximum impact

9:30 am Enabling Generalizable Models Through Smarter Data Aggregation & Validation Approaches

  • Neeraj Agrawal Director, Principal Scientist & Group Lead - Attribute Science Data Engineering, Amgen

Synopsis

  • Discuss the limitations of model generalizability due to inconsistent data protocols and lack of standardized evaluation
  • Review current industry efforts – such as consortia and cross-pharma collaborations, to pool and harmonize data
  • Learn how to avoid misleading conclusions by adopting statistically rigorous model performance assessments

NEW COMPANY!

10:00 am Implementing Active Learning Loops to Combat Bias & Improve Model Performance Over Time

  • Joel Karpiak Senior Director, Head of Protein Design & Informatics, GlaxoSmithKline

Synopsis

  • Discover how active and continuous learning systems allow models to improve with each new dataset, reducing initial bias
  • Examine the impact of bias from datasets favoring only “good” molecules and how this limits model creativity and exploration
  • Understand how integrating real-time lab results into training loops supports model refinement and validation

10:30 am Speed Networking

Synopsis

This session is a great opportunity to introduce yourself to the attendees that you would like to have more in-depth conversations with. This session is the ideal opportunity to get face-to-face time with many of the computational drug experts to establish meaningful business relationships.

11:00 am Morning Break

11:30 am Panel Discussion: The Data Dilemma – Standardizing Datasets to Improve Model Predictability

  • Kannan Sankar Senior Expert - Data Science & Bioinformatics, Novartis
  • Shipra Malhotra Senior Scientist - Biologics, Computational & Machine Learning, Takeda Pharmaceutical
  • Jenna Caldwell Associate Principal Scientist, AstraZeneca

Synopsis

  • What progress has been made in creating high-quality, standardized datasets to train robust predictive models?
  • How are industry leaders overcoming data silos and inconsistencies to drive better biologics insights?
  • What role do partnerships, consortia, or public databases play in improving interoperability and model performance?

Enabling Design of Complex Biologic Formats to Expand Therapeutic Possibilities

12:15 pm Expanding Predictive Design Beyond Monoclonal Antibodies to Bi- and Multispecific Protein Therapeutic

  • Norbert Furtmann Group Leader & Head of Artificial Intelligence Innovation Nanobody Platform, Computational & High-throughput Protein Engineering, Sanofi

Synopsis

  • Discuss challenges within multi-specific protein therapeutic design and discovery, exploring how computational tools can support the optimization tasks  
  • Learn about critical data gaps and strategic data generation approaches specifically tailored for multi-specifics
  • Examples and case studies demonstrating the application of computational and machine learning models to optimize next-generation multi-specific therapeutics

NEW COMPANY!

12:45 pm Designing & Predicting Properties of Emerging Biologic Formats Using AI & Structural Insight

Synopsis

  • Examine the current limitations of predictive tools for formats like VHHs, Fabs, ARC molecules, and RNA conjugates
  • Learn how AI can be tailored to support design and developability assessments across diverse biologic architectures
  • Encourage greater collaboration between structural biologists, AI developers, and discovery teams to accelerate innovation in this space

1:15 pm Lunch Break & Networking

2:15 pm BioDEVA: A Semi-Autonomous Data Pipeline to Build Antibody Developability Models

  • Michael Marlow Director, Biologics Research CMC, Boehringer Ingelheim

Synopsis

  • Design of a robust data pipeline that supports predictive modeling for key developability attributes
  • Tackle real-world challenges in sourcing and standardizing high-quality, diverse datasets
  • Enhance early decision-making so that time and effort are focused on the most promising molecules

2:45 pm Computational Approaches to Enable Challenge GPCR Agonist

Synopsis

  • Apply structure-based and AI-driven methods to identify and optimize novel GPCR agonists with improved specificity
  • Leverage predictive modeling and AI-based protein design to address limited GPCR structural data
  • Accelerate GPCR-targeted drug discovery by integrating virtual screening with ligand-receptor interaction analysis

3:15 pm Afternoon Break & Poster Session

Synopsis

Witness some of the latest research in the computational drug development space, presented by pharma, biotech, and solution providers in this spotlight poster session! Please visit the website for the T&Cs for presenting a poster.

3:45 pm Panel Discussion: From Hype to Impact – Realizing the Value of AI in Biologics Design

  • Gabor Oroszlan In Silico Lead, Project Manager, Senior Scientist, VRG Therapeutics
  • Jing Wang Director, Eli Lilly & Co.
  • Jinbo Xu Professor, Toyota Technological Institute at Chicago
  • Yao Fan Director - Protein Design & Optimization, Tectonic Therapeutic

Synopsis

  • What are the current limitations and tangible successes in applying AI/ML to biologics discovery and developability?
  • How are organizations measuring ROI from computational investments — speed, cost, accuracy, or all three?
  • What infrastructure, culture, and talent are needed to effectively scale AI adoption across discovery and development pipelines?

4:30 pm Benchmarking Paratope Binning to Improve Antibody Discovery

Synopsis

  • Prediction of the antibody paratope can guide antibody selection via paratope binning
  • Paratope binning shows a good correlation with binding profiles
  • Excitingly, paratope binning is less bound by sequence and expands clone discovery into a repertoire of sequence-diverse clones

5:00 pm Chair’s Closing Remarks

5:10 pm End of Conference Day One