A deep learning model trained on expressed transcripts across different tissue types reveals cell-type codon-optimization preferences
Nucleic Acids Research · 20.03.2025 · 53 (6)
DOI: 10.1093/nar/gkaf233 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nucleic Acids Research, online erschienen am 29.03.2025. Verfasst von Sandhiya Ravi und 6 weiteren Autor:innen. Beteiligte Einrichtung: Department of Genetic and Cellular Medicine, UMass Chan Medical School, Worcester, MA 01605, United States. Hauptthemen laut MEDLINE: Deep Learning, Codon Usage, Protein Biosynthesis, Codon. Die Arbeit wurde bisher 15-mal zitiert (Crossref) und nennt 81 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: NHLBI NIH HHS, NIAID NIH HHS, The Bassick Family Foundation.
Abstract (Originalsprache)
Species-specific differences in protein translation can affect the design of protein-based drugs. Consequently, efficient expression of recombinant proteins often requires codon optimization. Publicly available optimization tools do not always result in higher expression levels and can lead to protein misfolding and reduced expression. Here, we aimed to develop a novel deep learning (DL) tool using a recurrent neural network (RNN) to define cell type-dependent codon biases. Using gene expression data from three different tissue types (brain, liver, and muscle) and all secretory genes, we trained DL models to predict optimal codon usage. Codon-optimized sequences for test reporter genes exhibited enhanced protein expression compared to their original sequences and those optimized using a publicly available tool. Interestingly, DL models trained on genes expressed in liver cells (hepatocytes) resulted in the highest levels of expression when tested in vitro, irrespective of the cell type. Our findings also demonstrate that DL-based codon optimization algorithms can significantly enhance protein translation, particularly for secretory proteins, which are crucial for therapeutic applications. This research represents a novel approach to codon optimization with broader implications for protein-based pharmaceuticals, vaccine manufacturing, gene therapy, and other recombinant DNA products.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- Department of Genetic and Cellular Medicine, UMass Chan Medical School, Worcester, MA 01605, United States.
- Department of Microbiology, UMass Chan Medical School, Worcester, MA 01605, United States.
- Li Weibo Institute of Rare Diseases Research, UMass Chan Medical School, Worcester, MA 01605, United States.
ORCID-Profile: Tapan Sharma, Jun Xie, Phillip W L Tai
Förderung
- NHLBI NIH HHS
- NIAID NIH HHS
- The Bassick Family Foundation
- NINDS NIH HHS
- NIH HHS
- UMass Chan Medical School
- BRIDGE Fund Award
- National Institutes of Health
Zitieren
APA
Ravi, S., Sharma, T., Yip, M., Yang, H., Xie, J., Gao, G., & Tai, P. W. L. (2025). A deep learning model trained on expressed transcripts across different tissue types reveals cell-type codon-optimization preferences. Nucleic Acids Research, 53(6). https://doi.org/10.1093/nar/gkaf233
Vancouver
Ravi S, Sharma T, Yip M, Yang H, Xie J, Gao G, et al. A deep learning model trained on expressed transcripts across different tissue types reveals cell-type codon-optimization preferences. Nucleic Acids Research. 2025;53(6). doi:10.1093/nar/gkaf233
BibTeX
@article{Ravi2025,
author = {Ravi, Sandhiya and Sharma, Tapan and Yip, Mitchell and Yang, Huiya and Xie, Jun and Gao, Guangping and Tai, Phillip W L},
title = {A deep learning model trained on expressed transcripts across different tissue types reveals cell-type codon-optimization preferences},
journal = {Nucleic Acids Research},
year = {2025},
volume = {53},
number = {6},
doi = {10.1093/nar/gkaf233}
}
EUVIMED