kernaLM™ is unveiled at the 2nd Annual Computational RNA Design & Delivery Summit
2nd Annual Computational RNA Desing and Delivery Summit June 25-26 2025, Boston, MA USA
Oral Presentation:
Title: “Designing Selectively Translated mRNA via Computational Biology and Machine Learning”
Session Title: Optimizing RNA Sequences by Incorporating AI/ML for Uncovering Half-Life, Immunogenicity & Expression to Accelerate Potent RNA Therapies Towards the Clinic
Date and Time: Wednesday, June 25, 2025, 2:30 – 3:00 p.m. ET
Presenting Author: Jeff Ozdemir PhD, Co-founder and Head of Platform
Jeff Ozdemir, PhD, Co-founder and Head of Platform at Kernal Biologics, on his talk at the 2nd Annual Computational RNA Design & Delivery Summit 2025 introduced kernaLM™, our proprietary large language model that predicts mRNA stability and translation efficiency.
This breakthrough represents the culmination of years of iterative research and development, building upon our earlier explorations with genetic algorithms, convolutional neural networks (CNNs) for sequence pattern recognition, long short-term memory (LSTM) networks for capturing dependencies in mRNA folding dynamics, and capsule networks (CapsNets) for hierarchical feature representation in molecular structures. Each architectural evolution has contributed critical insights that ultimately informed the sophisticated transformer-based approach underlying kernaLM™, demonstrating our commitment to advancing the computational biology field through systematic innovation.
Also, many thanks to our SAB members Umut Eser, Co-founder and Chief AI Office at Weave, and Nikolai Slavov, Director of Parallel Squared Technology Institute and a Professor at Northeastern University, for their invaluable insights and guidance.