iProm-Sigma54: A CNN Base Prediction Tool for σ⁵⁴ Promoters

Area of Science:

• Molecular Biology
• Bioinformatics
• Genetics

Background:

• Sigma (σ) factors are essential for RNA holoenzymes to recognize and bind promoter regions in prokaryotic gene transcription.
• σ54 promoters are involved in diverse cellular processes and environmental responses, necessitating precise identification for gene regulation studies.

Purpose of the Study:

• To develop and validate a computational tool for accurate prediction of σ54 promoter sequences.
• To improve the understanding of prokaryotic gene regulation mechanisms involving σ54 promoters.

Main Methods:

• Development of a convolutional neural network (CNN) model named iProm-Sigma54.
• Utilizing a one-hot encoding scheme for sequence data input.
• Employing a five-fold cross-validation strategy and benchmark/test datasets for performance evaluation.

Main Results:

• The iProm-Sigma54 tool demonstrated higher prediction accuracy compared to existing methods for identifying σ54 promoters.
• Rigorous assessment using four metrics confirmed the tool's effectiveness.
• A publicly accessible web server was created for broader accessibility.

Conclusions:

• iProm-Sigma54 offers a robust and accurate method for σ54 promoter prediction.
• The tool facilitates research into prokaryotic gene regulation and environmental responses.
• The web server enhances the utility and application of this predictive model.