EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Transfer learning with pre-trained language models for protein expression level prediction in Escherichia coli

Chunhe Yang, YuLing Zhao, Ruoyu Wang, Haoran Li, Xiaoping Liao, Hongwu Ma

Synthetic and Systems Biotechnology · 2026-06 · 12 · 115-123

DOI: 10.1016/j.synbio.2025.11.012 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Synthetic and Systems Biotechnology, 2026-06. Verfasst von Chunhe Yang und 5 weiteren Autor:innen. Beteiligte Einrichtung: Biodesign Center, Key Laboratory of Engineering Biology for Low-carbon Manufacturing, Tianjin Institute of Industrial Biotechnology, Chinese Academy …. Die Arbeit wurde bisher 4-mal zitiert (Crossref) und nennt 32 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Chinese Academy of Sciences, National Natural Science Foundation of China, Innovation Fund.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Accurately predicting recombinant protein expression in Escherichia coli remains a long-standing challenge due to the multifactorial nature of gene regulation and translation. Existing computational approaches typically emphasize either codon usage or protein sequence features, limiting predictive accuracy and generalizability. Here we present TLCP-EPE, a transfer learning framework that, for the first time, fuses codon- and protein-level pre-trained language models to jointly capture determinants of expression. By fine-tuning CaLM and ProtT5 with low-rank adaptation (LoRA) and integrating their embeddings through a BiGRU-MLP predictor, TLCP-EPE learns expression-aware representations that outperform state-of-the-art methods. Across two independent test datasets, TLCP-EPE achieved robust performance (AUC 0.835 on codon data; AUC 0.713 on protein data), consistently surpassing conventional codon-based metrics and deep learning baselines. Our results demonstrate that dual-modal modeling of codon and protein sequences enables more accurate and generalizable prediction of expression levels, providing a powerful foundation for rational protein design and biomanufacturing applications.

Schlagwörter

Stichwörter der Autor:innen

Escherichia coliFused sequence representationsLoRA fine-tuningPre-trained language modelProtein expression predictionTransfer learning

Autor:innen und Institutionen

  • Biodesign Center, Key Laboratory of Engineering Biology for Low-carbon Manufacturing, Tianjin Institute of Industrial Biotechnology, Chinese Academy of Sciences, Tianjin, 300308, China.
  • Haihe Laboratory of Synthetic Biology, 300308, Tianjin, China.
  • National Technology Innovation Center of Synthetic Biology, Tianjin, 300308, China.
  • University of Chinese Academy of Sciences, Beijing, 100049, China.

Förderung

  • Chinese Academy of Sciences
  • National Natural Science Foundation of China
  • Innovation Fund

Zitieren

APA

Yang, C., Zhao, Y., Wang, R., Li, H., Liao, X., & Ma, H. (2026). Transfer learning with pre-trained language models for protein expression level prediction in Escherichia coli. Synthetic and Systems Biotechnology, 12, 115-123. https://doi.org/10.1016/j.synbio.2025.11.012

Vancouver

Yang C, Zhao Y, Wang R, Li H, Liao X, Ma H. Transfer learning with pre-trained language models for protein expression level prediction in Escherichia coli. Synthetic and Systems Biotechnology. 2026;12:115-123. doi:10.1016/j.synbio.2025.11.012

BibTeX
@article{Yang2026,
  author = {Yang, Chunhe and Zhao, YuLing and Wang, Ruoyu and Li, Haoran and Liao, Xiaoping and Ma, Hongwu},
  title = {Transfer learning with pre-trained language models for protein expression level prediction in Escherichia coli},
  journal = {Synthetic and Systems Biotechnology},
  year = {2026},
  volume = {12},
  pages = {115--123},
  doi = {10.1016/j.synbio.2025.11.012}
}
Kostenlos starten →