Related Experiment Video
Updated: Jul 12, 2026
05:49
Protein Engineering by Yeast Surface Display
Published on: November 29, 2024
Enhancing machine learning-based binder design with high-throughput screening: A comparison of mRNA and yeast display
Zhiyuan Yao1, McGuire Metts2, Avery K Huber3
1Department of Pharmacology, University of North Carolina School of Medicine, Chapel Hill, North Carolina, USA.
Protein Science : a Publication of the Protein Society
|July 10, 2026
Summary
Machine learning accelerates protein design, but experimental validation lags. This study compares mRNA and yeast display for screening ML-designed miniprotein binders, finding mRNA display offers better library coverage and enrichment of high-affinity binders.
Area of Science:
- Biochemistry
- Computational Biology
- Molecular Biology
Background:
- Machine learning (ML) enables rapid in silico design of miniprotein binders.
- Experimental validation methods are a bottleneck in protein binder discovery pipelines.
- Comparing in vitro display technologies for screening ML-designed libraries is crucial.
Purpose of the Study:
- To compare the performance of mRNA display and yeast surface display for screening ML-designed miniprotein binders.
- To evaluate the ability of each platform to identify functional binders targeting TLT-1 and B7-H3.
- To assess the impact of display technology on binder selection and library coverage.
Main Methods:
- Applied mRNA display and yeast surface display to screen a shared DNA library of ML-designed miniprotein binders.
- Screened 2009 designs against TLT-1 and 3159 designs against B7-H3.
- Performed biophysical characterization of selected binders.
Main Results:
- Both mRNA and yeast display identified functional binders.
- mRNA display preferentially enriched binders with slower dissociation rates.
- mRNA display achieved higher library coverage, potentially rescuing functional designs missed by yeast display.
- Selected binders exhibited strong binding affinities and high thermal stabilities.
Conclusions:
- Integrating ML-based protein design with rapid in vitro selection technologies like mRNA display provides a scalable framework for therapeutic miniprotein discovery.
- mRNA display offers advantages over yeast display in terms of library coverage and enrichment of binders with desirable kinetic properties.
- This integrated approach accelerates the discovery of high-quality therapeutic protein binders.
