8:00 am Check-In Opens & Light Breakfast
8:50 am Chair’s Opening Remarks
9:00 am Workshop: Deep Diving into the Fundamentals of Applying Computational Tools to RNA Design for Achieving Stable & Highly Expressive RNA Sequences
Synopsis
This 101 workshop will dive into the key principles and tools in using computational methodologies for RNA design, unveiling a comprehensive overview of how AI/ML approaches can be used to design, optimize, and analyze RNA sequences and their structures for various applications, including for protein expression, gene regulation and gene/RNA editing.
- Exploring computational tools which have been developed for codon optimization and secondary structure prediction
- Discovering how integrating new, high-quality datasets enhances the power of classic RNA computational approaches
- Diving into the critical role AI/ML plays in predicting RNA stability and interactions, and see how these advancements are revolutionizing RNA-based therapeutic design
10:45 am Morning Break & Speed Networking
Synopsis
This session is your opportunity to get face-to-face with many of the brightest minds working in the computational RNA field, and establish meaningful business relationships to pursue for the rest of the conference
Optimizing mRNA Sequences by Incorporating AI/ML for Uncovering Half-Life, Immunogenicity & Expression to Accelerate Potent RNA Therapies Towards the Clinic
11:45 am Advancing mRNA Sequence Design with Novel Embeddings & High- Throughput Integration
Synopsis
- Evaluating the accuracy of decoding to obtain mRNA sequences from different embedding strategies
- Testing innovative and more efficient embedding strategies, not based on language models, to represent and guide the search for improved mRNA sequences
- Discussing the integration of embedding and decoding strategies with high-throughput testing for various mRNA properties
12:15 pm Innovating RNA Therapeutics Design & Validation Through Sequencing
Synopsis
- Multidimensional datasets provide critical training data to improve potency and stability
- Validation of AI models with next-generation sequencing assays
- Data-driven optimizations to improve therapeutic efficacy
12:25 pm Leveraging a Small Language Model to Achieve Full-Length mRNA Analysis with Improved Prediction of Stability & Translation Efficiency
Synopsis
- CodonBERT is pre-trained on 10 million mRNA sequences to capture codon dependencies and interactions across different species
- mRNA-LM extends CodonBERT from the CDS to full-length mRNA by integrating two pre-trained BERT models on UTR regions
- Both CodonBERT and mRNA-LM show significant improvements in predicting mRNA properties compared to existing methods
12:55 pm Lunch & Networking
2:00 pm Developing a 5’UTR Large Language Model for Enhancing mRNA Function Predictions & Translation Efficiency
Synopsis
- Uncovering a large language model trained with 5’UTR regions and structural data for mRNA structure-function predictions
- Showcasing a large language model for improving prediction of translation efficiency, ribosome loading and mRNA expression levels to generate better mRNA medicines
- Optimizing a UTR library with optimized translation efficiency over traditional UTR designs
2:30 pm Applying Machine Learning to mRNA Sequence Design & LNP Development for Achieving Cell-Specific Delivery & Selective Translation
Synopsis
- Leveraging machine learning for design cell selective mRNA sequences to ensure precise targeting and selectivity in the right tissues
- Refining ionizable lipid discovery with AI to advance LNP formulations with optimal tissue tropism and biodistribution
- Delving into the future potential and applications of mRNA 2.0 in oncology and immunology
3:00 pm Afternoon Networking Break & Scientific Poster Session
Harnessing Experiment Design & Public Data Repositories fir Building High-Quality Datasets to Improve Reliability of Machine Learning Algorithms
3:45 pm Panel Discussion: Identifying Strategies for Data Sharing or Smaller Model Development to Improve Cost Feasibility of AI/ML Models for RNA & Lipid Design
Synopsis
- Discussing the potential of collaboration and data sharing between companies to accelerate data access for optimizing computational algorithms
- Evaluating the true impact of data quantity on computational model accuracy and prediction capability
- How to improve the cost efficiency of developing computational models for RNA design to make them more accessible
4:30 pm Combining ML-based Models with Sequence & Structure-Based Features for mRNA Optimization
Synopsis
- Incorporation of multiple state-of-the-art open-source ML models (autoregressive and autoencoding) with structural, sequence, and manufacturing heuristics into a coordinated framework to generate improved mRNA candidates
- Development of a rigorous data curation strategy, involving meticulous filtering and integration of publicly available datasets to precisely align with the objectives of our optimization challenge
- Validation and refinement of our end-to-end pipeline’s performance through in-house experimental data