Improving the generalization of protein expression models with mechanistic sequence information
Nucleic Acids Research · 24.01.2025 · 53 (3)
DOI: 10.1093/nar/gkaf020 ↗Kurz gefasst
Fachartikel in der Zeitschrift Nucleic Acids Research, online erschienen am 28.01.2025. Verfasst von Yuxin Shen und 2 weiteren Autor:innen. Beteiligte Einrichtung: School of Biological Sciences, University of Edinburgh, Edinburgh, EH9 3JH, United Kingdom. Hauptthemen laut MEDLINE: Gene Expression, Models, Genetic. Die Arbeit wurde bisher 11-mal zitiert (Crossref) und nennt 62 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0). Gefördert von: Biotechnology and Biological Sciences Research Council, United Kingdom Research and Innovation, UK Medical Research Council.
Abstract (Originalsprache)
The growing demand for biological products drives many efforts to maximize expression of heterologous proteins. Advances in high-throughput sequencing can produce data suitable for building sequence-to-expression models with machine learning. The most accurate models have been trained on one-hot encodings, a mechanism-agnostic representation of nucleotide sequences. Moreover, studies have consistently shown that training on mechanistic sequence features leads to much poorer predictions, even with features that are known to correlate with expression, such as DNA sequence motifs, codon usage, or properties of mRNA secondary structures. However, despite their excellent local accuracy, current sequence-to-expression models can fail to generalize predictions far away from the training data. Through a comparative study across datasets in Escherichia coli and Saccharomyces cerevisiae, here we show that mechanistic sequence features can provide gains on model generalization, and thus improve their utility for predictive sequence design. We explore several strategies to integrate one-hot encodings and mechanistic features into a single predictive model, including feature stacking, ensemble model stacking, and geometric stacking, a novel architecture based on graph convolutional neural networks. Our work casts new light on mechanistic sequence features, underscoring the importance of domain-knowledge and feature engineering for accurate prediction of protein expression levels.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Substanzen
Autor:innen und Institutionen
- School of Biological Sciences, University of Edinburgh, Edinburgh, EH9 3JH, United Kingdom.
- Institute for Genetics and Cancer, University of Edinburgh, Edinburgh, EH4 2XU, United Kingdom.
- School of Informatics, University of Edinburgh, Edinburgh, EH8 9AB, United Kingdom.
ORCID-Profile: Grzegorz Kudla, Diego A Oyarzún
Förderung
- Biotechnology and Biological Sciences Research Council
- United Kingdom Research and Innovation
- UK Medical Research Council
- Wellcome Trust
- UKRI Centre for Doctoral Training in Biomedical AI
Zitieren
APA
Shen, Y., Kudla, G., & Oyarzún, D. A. (2025). Improving the generalization of protein expression models with mechanistic sequence information. Nucleic Acids Research, 53(3). https://doi.org/10.1093/nar/gkaf020
Vancouver
Shen Y, Kudla G, Oyarzún DA. Improving the generalization of protein expression models with mechanistic sequence information. Nucleic Acids Research. 2025;53(3). doi:10.1093/nar/gkaf020
BibTeX
@article{Shen2025,
author = {Shen, Yuxin and Kudla, Grzegorz and Oyarzún, Diego A},
title = {Improving the generalization of protein expression models with mechanistic sequence information},
journal = {Nucleic Acids Research},
year = {2025},
volume = {53},
number = {3},
doi = {10.1093/nar/gkaf020}
}
EUVIMED