Core Perturbomes of Escherichia coli and Staphylococcus aureus Using a Machine Learning Approach.

Area of Science:

• Microbiology and Systems Biology
• Computational Biology and Bioinformatics

Background:

• The core perturbome represents a conserved molecular network essential for maintaining homeostasis under stress.
• Escherichia coli and Staphylococcus aureus are clinically relevant prokaryotic models with significant biological impact.

Purpose of the Study:

• To identify and functionally characterize the core perturbome in Escherichia coli and Staphylococcus aureus using machine learning.
• To discover core molecular signatures that distinguish between control and perturbed conditions in these bacteria.

Main Methods:

• Utilized gene expression data from E. coli (132 samples) and S. aureus (156 samples).
• Applied machine learning algorithms (KNN, RF, SVM) for feature selection and classification to identify core genes.
• Performed molecular interaction and functional enrichment analyses on identified gene sets.

Main Results:

• Identified a core perturbome of 55 genes (9 hubs) in E. coli and 46 genes (8 hubs) in S. aureus.
• Achieved high classification accuracies (82.6% for E. coli, 85.1% for S. aureus) in distinguishing conditions.
• Characterized core genes associated with metabolism, synthesis/degradation, transcription regulation, and virulence.

Conclusions:

• The identified core perturbomes provide insights into bacterial stress response mechanisms.
• These findings highlight potential therapeutic targets and biomarkers for bacterial stress and infection.