Using complete genome comparisons to identify sequences whose presence accurately predicts clinically important...

Area of Science:

• Microbiology
• Genomics
• Bioinformatics

Background:

• Distinguishing pathogenic bacterial strains, such as enterohemorrhagic Escherichia coli (EHEC), from harmless ones is crucial for clinical settings.
• Current methods for determining pathogenic phenotypes are often time-consuming and costly.

Purpose of the Study:

• To develop a simple, rapid, and inexpensive method for predicting pathogenic bacterial phenotypes.
• To identify specific DNA segments that can accurately predict the pathogenicity of bacterial isolates.

Main Methods:

• A novel approach comparing complete bacterial genomes to identify short homologous DNA segments predictive of phenotypes.
• Utilizes binary string alignment of genomes without requiring full genome alignments.
• Clinical application involves PCR amplification of identified predictive DNA segments.

Main Results:

• The method successfully identified EHEC strains of E. coli and distinguished E. coli from Shigella in silico and experimentally.
• In silico analysis showed that using as few as 8 predictive sequences, with amplification of just three, yields a >0.99 probability of correct identification.
• Experimental screening of 98 isolates correctly identified all strains as E. coli/Shigella or EHEC/non-EHEC E. coli.

Conclusions:

• The proposed DNA segment-based method offers a robust, rapid, and cost-effective approach for predicting bacterial pathogenicity.
• The technique is highly accurate and resilient to occasional PCR amplification failures, making it suitable for clinical diagnostics.