December 5-7, 2017
Boston, MA

Day One
Wednesday, December 6 2017

Day Two
Thursday, December 7 2017

In Silico Modeling for Antibody Engineering & Risk Mitigation

08.30
Balancing Coarse-Grained & Molecular Models to Predict Low & High Concentration Protein-Protein Interactions

Synopsis

• Assessing different levels of coarse-grained molecular models to predict attractive vs. repulsive protein-protein interactions at low and high protein concentrations
• Balancing the level of structural resolution with the computational burden for predictions of protein-protein interactions
• Methods to identify key domain-domain interactions or amino acid pairings that cause strong attractions that promote protein aggregation

09.00
Predictive Tools for Developability Assessment of Antibody Molecules

Synopsis

• Introducing biopharmaceutical informatics, an approach to predict antibody solution properties using standardized experiments, molecular modeling, and machine learning
• Describing all the experiments performed and discuss how experimental properties are related to each other
• Presenting equations to predict protein-protein interactions from molecular modeling and machine learning

09.30
Morning Refreshments & Networking

10.30
Roundtable Discussions Followed by Moderator Feedback & Audience Debate

Synopsis

This 2 hour session offers the opportunity to catch up on the latest advancements within the field and learn from your fellow colleagues in this interactive and informal format. A valuable chance for attendees to unite around hot topics and debate best practice. No more sitting quietly, this is a dedicated opportunity for you to voice your experiences and identify unique solutions.

  1. Informing computational models through insights from bioinformatics.
  2. Balancing Model Complexity Against Function.
  3. Formulation design for biologics in the age of lab automation and biological performance screening.
  4. Predicting biophysical characterisation to enhance candidate selection and lead optimisation

12.30
Lunch & Networking

Advanced Molecular Assessment Tools for Antibody Design

13.30
Enabling Developability via Molecular Modeling

Synopsis

• Reviewing key molecular assessment developability tools, their scope, and validation
• Exploring cutting-edge approaches for formulation analysis and design
• Highlighting strategies for implementation of developability

14.00
Machine Learning for Predicting Developability from Sequence

Synopsis

• Discussing experimental data to validate and motivate in silico model development
• Applying algorithm to predict experimental data
• Utilizing HTP methods to design and prioritize antibody libraries

14.30
Afternoon Refreshments & Networking

Applying MAb Learnings to the Developability of Other Biologic Modalities

15.00
Predicting Biophysical & Drug-Like Properties of Bi/Multispecific Molecules

  • Feng Dong Senior Scientist, Abbvie BioResearch Center

Synopsis

• Using rationale-design to enable manipulation of agonist function
• DVD-IG: overcoming issues in combining two or more variabile domains
• Highlighting the necessity of stability and interface engineering

15.30
Computational Modeling of Precision Bacteriotherapies

  • Georg K. Gerber Assistant Professor, Harvard Medical School, Co-Director, MA Host- Microbiome Center

Synopsis

• Showcasing what precision bacteriotherapies are and articulating the challenges in designing and predicting the effects of such therapies
• Providing a conceptual understanding of dynamic systems based modeling for bacteriotherapy applications
• Develop a conceptual understanding of approaches for inferring dynamical systems model parameters from in vivo microbiome data

16.00
Identifying Global Trends Across the RNA Binding Protein (RBP) Interactome Using Deep Learning

  • Varun Shivashankar Senior Scientific Computing Engineer, Novartis Institutes for Biomedical Research

Synopsis

• Integrating RNA structure and genome sequence knowledge of cognate binding elements with our eCLIP data to have a better understand of binding dynamics
• Leveraging this learning model in predicting RBP binding site on a genomic scale and porting this to cell lines without eCLIP data
• Expanding scope of model to predict other interaction fingerprints

16.30
Close of Conference