AI Insight
Researchers developed a new method called Latent-Space Codon Optimization (LSCO) that improves the selection of synonymous codons to maximize protein expression in cells. The approach converts the discrete codon selection problem into a continuous optimization problem by using a pretrained mRNA language model, allowing efficient gradient-based search guided by a data-driven expression predictor. Testing on antibody expression data showed LSCO outperformed both traditional frequency-based methods and modern deep learning baselines while maintaining appropriate biophysical properties.
Why it matters
This advance could significantly improve therapeutic protein production and mRNA vaccine development by providing a more effective way to optimize genetic sequences for higher protein yields. The method addresses a fundamental bottleneck in biotechnology where current optimization tools rely on heuristic proxies that poorly predict actual expression levels.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Abstract: Codon optimization, the process of selecting synonymous codons to improve mRNA translation efficiency and protein expression, is central to therapeutic protein production and mRNA vaccines, yet it remains a hard problem. The design space is discrete and combinatorially large, precluding gradient-based methods, and existing tools rely on heuristic proxies (e.g., Codon Adaptation Index or GC-content) that poorly capture true expression. We introduce Latent-Space Codon Optimization (LSCO), which recasts this discrete problem as a continuous one by mapping sequences into the latent space of a pretrained mRNA language model, enabling efficient gradient-based search. LSCO combines four components: a data-driven expression objective from an uncertainty-aware predictor, a Minimum-Free-Energy regularizer for structural stability, a naturalness prior from a protein-to-codon back-translation model, and constrained decoding for protein fidelity. On a real-world, wet-lab antibody expression dataset, LSCO outperforms simple frequency-based, as well as modern deep generative baselines in predicted expression, while retaining suitable biophysical properties.
Source: Predictor-Guided Latent Space Codon Optimization for Maximizing Protein Expression