An Automated Model Test System for Systematic Development and Improvement of Gene Expression Models

Area of Science:

• Synthetic biology
• Computational biology
• Biophysics

Background:

• Accurate gene expression models are crucial for engineering synthetic biology systems.
• Current models face challenges in prediction accuracy, limiting design efficiency.
• Reducing trial-and-error experimentation is key for advancing synthetic biology.

Purpose of the Study:

• To develop an automated model test system for evaluating gene expression models.
• To introduce a new metric, model capacity, for robust cross-dataset comparisons.
• To identify areas for improvement in existing gene expression models.

Main Methods:

• Combined advanced statistics and machine learning with a database of 9862 genetic systems.
• Developed an automated system to quantify model accuracies and evaluate mechanistic hypotheses.
• Introduced and utilized the information theoretic metric 'model capacity'.

Main Results:

• Compared six translation initiation rate models and evaluated 100 mechanistic hypotheses.
• Uncovered novel sequence determinants influencing protein expression levels.
• Developed an improved biophysical model for translation initiation rate with enhanced accuracy.

Conclusions:

• Automated model test systems significantly accelerate the development of gene expression models.
• This work advances synthetic biology towards a mature engineering discipline.
• The developed system facilitates accurate prediction of sequence-function relationships.