Explore the Agenda
8:00 am Check-In & Morning Coffee
8:50 am Chair’s Opening Remarks
Leveraging Computational & AI/ML Approaches to Predict & Optimize Leads for Developability Parameters & Identify Liabilities Ahead of CMC & Formulation
9:00 am Integrating Structure‑Based & Data‑Driven Approaches for Antibody Property Prediction
- State of the art for sequence based approaches for antibody properties: performance of antibody LLMs
- Value added by structure-based physics informed descriptors
- Embedding property prediction models into biologics discovery workflows
9:30 am Unlocking Protein Insights via Seamless Access to AI Models at Scale
- AI breakthroughs are revolutionizing drug discovery and molecular biology
- Research teams face practical barriers including fragmented toolsets, inconsistent infrastructure across cloud and on-premises environments, and complex interfaces hinder adoption by non-computational scientists
- In competitive drug discovery, speed to insight drives success
10:00 am Benchmarking Antibody-Antigen Affinity Prediction Across Sequence-, Structure-, & Physics-Based Methods
- Structure-aware approaches outperform sequence-based models, highlighting the importance of explicit 3D epitope–paratope interactions
- PLM and LLM methods underperform overall, indicating limits of sequence-only representations for binding affinity prediction
- Reliable prediction remains driven by structural and physicochemical information, with hybrid approaches showing the most promise
10:30 am From the Data Gap to Foundational Models: Standardized, Multi-Format Data for Predictive Antibody Developability
- Overviewing Ginkgo’s PROPHET-Ab platform for generating high-throughput, standardized developability data across IgGs, VHHs, and bispecifics for partners
- Highlighting publicly available ML-ready developability datasets, including recent learnings from the rational design and analysis of a 160-member bsAb library
- Discussing projects for generating the large amounts of purpose-built data and the impacts of those large datasets on predictive models
10:40 am Morning Break & Networking
11:00 am Overcoming Challenges in Drug Discovery with Lilly TuneLab
- TuneLab provides access to Lilly-trained AI models to help accelerate breakthrough medicines to patients
- The platform employs federated learning, a privacy-preserving approach that enables biotech developers to tap into Lilly’s AI models without directly exposing their proprietary data
- Hear real-world examples from partners using TuneLab to accelerate their drug discovery work
Translating & Scaling Findings to Develop AI/ML Models for More Complex Biologics
Including Bispecifics, Multispecifics & Conjugated Drug Products
11:30 am Multi-Modal and Multi-Objective Optimization for Functional and Developable Antibody Design
- Computational drug discovery is a multi-layered challenge spanning many platforms and modalities; needing a scalable, integrated decision system. Value comes not just from accurate predictive oracles, but from the strategy for using them.
- The core bottleneck is data fit and quality, not volume. Both the lead optimization and developability prediction cases point to the need for data tailored to the specific projects. Models must not only predict accurately, but stay grounded in the constraints of their use case, and be able to guide new data generation that improves predictive performance over time.
- Automation infrastructure, i.e., design automation, is what turns good models into real acceleration. Without robust data/model pipelines and workflow automation linking generative models, predictive oracles, and lab feedback, even accurate individual models fail to translate into faster discovery.
12:00 pm Closed Loop Antibody Engineering: Combining High-Throughput Campaigns with ML for Multi-Objective Design
- Closed loop to static screening: Antibody discovery becomes dramatically more efficient when HTx data and ML are integrated into an iterative, learning system, not a one-off screening pipeline
- Multi-objective is the real bottleneck: Optimizing affinity alone is solved; the challenge (and opportunity) is balancing affinity, specificity, and developability simultaneously, and this requires new modeling and data strategies
- Data quality beats model complexity: The biggest gains come from designing the right experiments (diverse, informative, comparable), not from using more complex ML models
12:30 pm Designed to Spec: ConvergeAB for Multi-Parameter Antibody Engineering
- Discover how ConvergeAB engineers multiple antibody properties such as humanness, immunogenicity, thermal stability, affinity, and selectivity simultaneously by tuning each parameter to a pre-defined target profile
- Walk through real-world case studies pushing the boundaries of what single-pass AI design can achieve: engineering a biobetter of Cetuximab, building species cross-reactivity from scratch, and executing selective affinity maturation against a challenging target
- Learn how ConvergeAB delivers these capabilities through an intuitive, secure platform with built-in compute, enterprise-grade data protection, and workflows designed for all scientists
12:40 pm Lunch Break & Networking
1:30 pm Harnessing AI for Smarter Antigen Selection & More Effective TCE Design
- Using ML to integrate diverse datasets and map antigen prevalence across cancer and normal tissues
- Applying the 3T-TRACE platform to assess TCR and TCR-mimetic specificity and off-target cross-reactivity
- Engineering highly specific and safe T-cell engagers against shared targets for effective treatment of solid tumors
2:00 pm Application of AI/ML for Predicting CIEX Binding Differences of mAb Building Blocks
- Sequence pI, the most widely used CIEX binding predictor is a poor predictor of binding to CIEX
- LLM descriptors can be used to predict aCIEX retention time
- 100s of aCIEX data points and their sequence, structure and pLLM based features can be leveraged for selecting and/or designing mAb building blocks
2:30 pm Afternoon Networking & Refreshment Break
Supercharging Molecular Dynamics Approaches with AI/ML Models to Reflect the Folding
Energetics of Biologic Therapeutics
3:00 pm Multiscale Molecular Modeling of mAbs: From Structure to Attributes
- Multiscale modeling enables us to think beyond single structure-based predictions for mAbs
- Bottom-up molecular dynamics approaches could be the key tool that connects molecular structure to its site-specific and bulk attributes
3:30 pm Structure-Guided Computational Insights into Antibody-Drug Conjugates Using Molecular Dynamics Simulations
- In silico predictive modeling of ADC properties
- Importance of MD for ADC design
- ADC design considerations for stability
In-Conference Workshop
4:00 pm Lab Automation for AI-Enabled Biologic Therapeutics Design: Closing the Loop Between Compute, Wet-Lab Experimentation & Iterative AI/ML Model Learning
AI/ML models are only as powerful as the experimental wet-lab insights that are used as input data to generate feedback. This workshop explores how lab automation, assay standardization, and closed‑loop experimentation are transforming biologic therapeutic discovery by enabling continuous, self‑improving AI workflows.
Participants will explore:
- The role of lab automation in enabling AI‑driven protein discovery at scale
- How to design lab workflows that generate high‑quality, AI‑usable experimental data rather than isolated results
- How to integrate automated wet‑lab experimentation with in silico prediction pipelines to create continuous lab‑in‑the‑loop learning cycles
- Where lab automation delivers the greatest ROI today, such as, binder screening and affinity maturation, developability profiling and high‑throughput functional and biophysical assays
- Variability across automated platforms, assays, and labs, unpacking the implications for AI/ML model performance
- Human‑in‑the‑loop decision‑making, QA and QC, where expert input remains critical despite automation
- Scaling automated discovery workflows while maintaining data reproducibility, QC, and regulatory confidence