Can protein expression be ‘solved’?

Catherine Baranowski (Author), Hector Garcia Martin (Author), Diego A. Oyarzún (Author), Aviv Spinner (Author), Bijoy Desai (Author), Christopher J. Petzold (Author), Evangelos-Marios Nikolados (Author), Sebastian Jaaks-Kraatz (Author), Aljaž Gaber (Author), Robert J. Chalkley (Author), Devin Scannell (Author), Rachel Sevey (Author), Michael C. Jewett (Author), Peter J. Kelly (Author), Erika A. DeBenedictis (Author)

Abstract

Heterologous protein expression is a fundamental technique used frequently in modern day biology. It enables scientific exploration of protein function as well as development of lifesaving medicines and economically impactful industrial products. Protein expression experiments primarily remain an experience-guided trial and error situation, even though it is an approach used by nearly all biologists. Generating an openly available, large-scale protein expression dataset that spans organisms and uses a standard experimental approach would provide the machine learning community with a foundation for building a multispecies predictive model of expression. A predictive model of protein expression would have a profound commercial impact and could replace countless hours of experimentation with a higher-probability directed approach.

Keywords

izražanje proteinov;strojno učenje;napovedni modeli;odprte baze podatkov;protein expression;machine learning;predictive models;open datasets;

Data

Language:	English
Year of publishing:	2025
Typology:	1.02 - Review Article
Organization:	UL FKKT - Faculty of Chemistry and Chemical Technology
UDC:	577.112:004.85
COBISS:	238521091
ISSN:	0167-7799
Views:	176
Downloads:	26
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	Slovenian
Secondary keywords:	izražanje proteinov;strojno učenje;napovedni modeli;odprte baze podatkov;
Type (COBISS):	Article
Pages:	str. 1-19
Volume:	ǂVol. ǂ
Issue:	ǂiss. ǂ
Chronology:	2025
DOI:	10.1016/j.tibtech.2025.04.021
ID:	26558431