Conference Day Two
Day Two
Thursday 26th October 2023
8:20 am Chair’s Opening Remarks
Predictability for Biologics Developability & Understanding Toxicity
8:30 am Leveraging AI-ML Approaches in Biologics Discovery & Process Development
Synopsis
- Introducing Teva Pharmaceuticals and their work
- Applying AI-ML tools in target identification for biologics
- Ensuring the optimal antibody process development by implementing AI-ML tools
9:00 am Exploring Computational Methods to Predict Biologics Safety
Synopsis
- Addressing the importance of the 3Rs to support improving translation into humans
- Exploring methods to overcome data shortages in a safety perspective
- Leveraging computational modelling tools to remove survivor bias
9:30 am Predicting Immunogenicity in Biologics Beyond the Clinic Through Next- Generation Deep Learning Tools
Synopsis
- Implementing omics, and data analysis method to train models for clinical candidate selection
- Leveraging a protein language model to predict immunogenicity among other properties
- Evaluating drug sensitivity of biologics candidates in the clinic using predictive models
10:00 am Morning Break & Networking Coffee
Preparing For Emerging Applications of Computational Approaches
10:30 am Mastermind session: Preparing for the Future: Rethinking Career Pathways & Attracting Next Gen Scientific Talent to Drive Growth in the Space
Synopsis
- Understanding the interplay between computational sciences and experimental platforms as the industry grows
- Assessing the kind of talent that needs to be nurtured
- Determining the kinds of infrastructure and the kinds of data required to empower this
11:00 am Surface ID: A Deep Learning-based Molecular Descriptor & a Useful Tool for Antibody Characteristics
Synopsis
- Explaining the importance of surface-based representation
- Developing a geometric deep learning model based on self-supervised learning
- Highlighting the applications of this model for surface similarity comparison, and antibody paratope-based clustering
11:30 am Predicting mAb Subcutaneous Bioavailability from In Silico Structural Features
Synopsis
- Assessing the current animal model gaps to test monoclonal antibody bioavailability
- Addressing the importance of understanding bioavailability for mAbs administered subcutaneously
- Leveraging in silico tools to predict bioavailability
12:00 pm Lunch Break
1:30 pm Breakout Roundtable Discussions:
Synopsis
This session is your opportunity to share your most pressing challenges, and work as a group to come up with solutions that you can implement right away! Each topic area will have several small groups, and each group will have 20 minutes to discuss ways to overcome barriers across biologics developability and computational methods. Groups will then share their findings with all attendees, giving you maximum exposure to new ideas!
1. Discussing the Realistic Limitations of AI
Synopsis
- Addressing the current excitement following the introduction of chatGPT
- Reiterating the importance of innovation whilst ensuring new computational approaches are validated
- Understanding the expectations of experimental and computational scientists with the current trends in the space
2. Highlighting Machine Learning Approaches to Overcome Developability Issues
Synopsis
- Exploring the current methods used to balance formulation with biologics properties
- Understanding the downstream implications of not overcoming these issues on HPC chromatography and more
1:00 pm Breakout Roundtable Discussions:
Synopsis
This session is your opportunity to share your most pressing challenges, and work as a group to come up with solutions that you can implement right away! Each topic area will have several small groups, and each group will have 30 minutes to discuss ways to overcome barriers across biologics developability and computational methods. Groups will then share their findings with all attendees, giving you maximum exposure to new ideas!
Emerging Integration of Computational Approaches with Biologics
1:45 pm Predicting Polyreactivity of Antibodies by Implementing Protein Language Models
Synopsis
- Understanding the applications of deep learning models to predict monospecific, bispecific, and heavy-chain-only antibody properties
- Exploring the applications of techniques to develop deep learning models to predict polyreactivity
- Uncovering the future hope for these models in antibody discovery
2:15 pm Merging Antibody Developability with De Novo Design
Synopsis
- Exploring the interplay between structure and sequence-based biologics design
- Implementing machine learning tools for structure-based de novo design
- Highlighting the data challenges that limit this
2:45 pm Applications & Limitations of Quantum Computing in Drug Discovery
Synopsis
- Addressing the current limitations in hardware to support this technology
- Understanding the rationale and strategies taken to implement this
- Unravelling the future hopes for this technology