# Screening and machine learning-based prediction of translation-enhancing peptides that reduce ribosomal stalling in Escherichia coli

## TL;DR

Researchers identified peptides that reduce ribosome stalling in E. coli and used machine learning to predict their effectiveness.

## Contribution

A machine learning model was developed to predict translation-enhancing peptides based on experimental screening.

## Key findings

- Translation-enhancing peptides were identified that reduce SecM AP-induced ribosomal stalling in E. coli.
- The fourth amino acid in tetrapeptides significantly influences their translation-enhancing activity.
- A random forest machine learning model accurately predicted TEP activity based on experimental data.

## Abstract

We previously reported that the nascent SKIK peptide enhances translation and alleviates ribosomal stalling caused by arrest peptides (APs) such as SecM and polyproline when positioned immediately upstream of the APs in both Escherichia coli in vivo and in vitro translation systems. In this study, we conducted a comprehensive screening of translation-enhancing peptides (TEPs) using a randomized artificial tetrapeptide library. The screening focused on the ability of the peptides to suppress SecM AP-induced translational stalling in E. coli cells. We identified TEPs exhibiting a range of translation-enhancing activities. In vitro translation analysis suggested that the fourth amino acid in the tetrapeptide influences the reduction of SecM AP-mediated stalling. Additionally, we developed a machine learning model using a random forest algorithm to predict TEP activity, which showed a strong correlation with experimentally measured activities. These findings provide a compact peptide toolkit and a data-driven approach for alleviating AP-induced ribosome stalling, with potential applications in synthetic biology.

Translation-enhancing peptides (TEPs) that reduce ribosomal stallling in Escherichia coli were identified through comprehensive screening, and their activity was predicted using a machine learning model.