EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 4.0

Language model-based self-training reduces labeled data requirements by 99% for biological sequence classification

Jingwen Liu, Danmo Gao, Yan Yuan, Ziying Wu, Weidong Li, Chao Wei

Briefings in Functional Genomics · 09.01.2026 · 25

DOI: 10.1093/bfgp/elag005 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Functional Genomics, online erschienen am 15.06.2026. Verfasst von Jingwen Liu und 5 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer Science and Artificial Intelligence, Hubei University of Technology, 28 Nanli Road, Hongshan District, Wuhan 430068, China. Hauptthemen laut MEDLINE: Language. Die Arbeit nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Hubei University of Technology.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The prediction of a biological sequence’s function is an exploding field that could prove to be an invaluable tool in deciphering molecular mechanisms of diseases and genetic variation. However, existing techniques suffer from a very low number of experimentally verified labels as well as the complexity of modeling sequence context. Previous work only focuses on semi-supervised learning (SSL) or pre-trained language models (PLMs) individually and does not consider their complementarity: whereas PLMs capture syntactic conservatism and long-range dependency from huge unlabeled sequence data as features, SSL updates label-level certainty via confidence-weighted pseudo-label selection. To realize integrating these benefits, we make a creative use of PLMs as strong feature extractors to make the sequence semantics extracted, and meanwhile leverage the SSL to constrain the decision boundary by selecting reliable pseudo-labels. We show that our framework achieves competitive performance compared to full supervision while using far fewer labeled samples, on two biological prediction tasks: DNA-binding protein (DBP) and non-coding RNA (ncRNA) detection. In general, our approach, developed here, not only offers an efficient solution for discovering new DBP and ncRNA in low-resource settings where experimental verification is infeasible, but also lays a solid methodological foundation for biological sequence classification in computational biology. Key points Unlabeled data significantly improves model accuracy in semi-supervised learning. The language model-based self-training method outperforms traditional semi-supervised approaches like TSVM. It achieves competitive performance against state-of-the-art fully supervised methods, requiring only 1% labeled data for effective classification and accurately identifying novel biomolecules.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

LanguageRNA, UntranslatedDNA-Binding ProteinsHumans

Stichwörter der Autor:innen

biological sequence classificationpre-trained language modelsself-training

Substanzen

RNA, UntranslatedDNA-Binding Proteins

Autor:innen und Institutionen

  • School of Computer Science and Artificial Intelligence, Hubei University of Technology, 28 Nanli Road, Hongshan District, Wuhan 430068, China.
  • Department of Information Sciences and Technology, Wen Hua College, 8 Wenhuayuan Road, East Lake High-tech Development Zone, Wuhan 430074, China.
  • Hubei Provincial Key Laboratory of Green Intelligent Computing Power Network, 28 Nanli Road, Hongshan District, Wuhan 430068, China.
  • Hubei Provincial Engineering Research Center for Digital & Intelligent Manufacturing Technologies and Applications, 28 Nanli Road, Hongshan District, Wuhan 430068, China.

Förderung

  • Hubei University of Technology

Zitieren

APA

Liu, J., Gao, D., Yuan, Y., Wu, Z., Li, W., & Wei, C. (2026). Language model-based self-training reduces labeled data requirements by 99% for biological sequence classification. Briefings in Functional Genomics, 25. https://doi.org/10.1093/bfgp/elag005

Vancouver

Liu J, Gao D, Yuan Y, Wu Z, Li W, Wei C. Language model-based self-training reduces labeled data requirements by 99% for biological sequence classification. Briefings in Functional Genomics. 2026;25. doi:10.1093/bfgp/elag005

BibTeX
@article{Liu2026,
  author = {Liu, Jingwen and Gao, Danmo and Yuan, Yan and Wu, Ziying and Li, Weidong and Wei, Chao},
  title = {Language model-based self-training reduces labeled data requirements by 99% for biological sequence classification},
  journal = {Briefings in Functional Genomics},
  year = {2026},
  volume = {25},
  doi = {10.1093/bfgp/elag005}
}
Kostenlos starten →