Assessing Codon Language Models for Context-Aware Codon Optimization in Nucleic Acid-Based Medicines
Codon optimization uses synonymous sequence changes to improve the expression and therapeutic performance of nucleic acid-based medicines. Masked language models (MLMs) have recently been proposed as alternatives to traditional, frequency-based codon optimization approaches, yet whether they offer a meaningful advantage over such simpler methods remains unclear. Here we benchmark three prominent…
Assessing Codon Language Models for Context-Aware Codon Optimization in Nucleic Acid-Based Medicines explores the potential benefits of using masked language models (MLMs) over traditional codon optimization methods for enhancing expression and therapeutic performance in nucleic acid-based medicines. Three prominent MLMs, CaLM, EnCodon, and CodonTransformer, were compared across backtranslation fidelity, sequence generation, and nine molecular phenotype prediction tasks.
Additionally, experimental evaluation using a secreted embryonic alkaline phosphatase (SEAP) reporter was conducted.
The findings revealed that the MLMs exhibited varying levels of amino-acid fidelity and generated unique synonymous sequence variants. However, no single model consistently outperformed the others across all benchmark tasks. Simple sequence features still demonstrated competitive performance in several settings. Interpretability analysis indicated that the MLMs incorporated a broader context of codons for making predictions, as opposed to frequency-based approaches.
In vitro data demonstrated that MLM-designed sequences outperformed both conventional and commercially derived sequences in terms of transient and stably integrated expression. This performance improvement suggests that MLMs can capture translational context beyond codon frequency, establishing themselves as effective and complementary tools for codon optimization. Furthermore, the study suggests that sampling across multiple MLMs may increase the likelihood of identifying high-performing therapeutic sequences.
Written by urgent.news from bioRxiv's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
