Multi-Objective-Guided Generative Design of mRNA with Therapeutic Properties

ICML 2025 Workshop FM4LS Submission60 Authors

Keywords: mrna, diffusion, design, therapeutics

TL;DR: Multi-objective-guided diffusion for generation of mRNA transcripts optimized across several therapeutic properties.

Abstract: Therapeutic mRNA design requires simultaneous optimization of multiple competing properties, including stability, translation efficiency, and protein expression—a challenge that current single objective approaches struggle to address systematically. We introduce mRNA generation via Unrolled Trajectories and Informed Latent UpdateS (mRNAutilus), the first multi-objective guided generative model for simultaneous mRNA codon optimization and de novo design of untranslated region (UTR) sequences. mRNAutilus combines a Masked Diffusion Model (MDM) trained on 6 million evolutionary-scale mRNA sequences with Monte Carlo Tree Search to guide generation toward Pareto-optimal solutions across multiple therapeutic properties. As objective functions, we use the latent mRNAutilus embeddings to train classifiers for predicting half-life, ribosome profiling, and translation rate. Guided by our classifiers, we demonstrate that mRNAutilus successfully generates diverse, high-quality mRNA sequences with enhanced properties compared to natural sequences while preserving coding regions for therapeutic applications, establishing a new paradigm for rational mRNA design.

Submission Number: 60