Evaluating the Reliability of eBURST in Bacterial Population Analysis
Author Information
Author(s): Katherine ME Turner, William P Hanage, Christophe Fraser, Thomas R Connor, Brian G Spratt
Primary Institution: Imperial College, St. Mary's Hospital Campus, London, UK
Hypothesis
How reliable is the eBURST program for identifying groups of closely related bacterial strains?
Conclusion
The study found that eBURST is generally reliable for identifying clonal complexes in bacterial populations, but its performance declines with high rates of recombination.
Supporting Evidence
- eBURST accurately identified 90-98% of true ancestor-descendant relationships in strictly clonal simulations.
- Performance declined with increasing recombination to mutation ratios, dropping to 61% accuracy at high ratios.
- Most bacterial species analyzed fell within the reliable performance range of eBURST.
Takeaway
eBURST helps scientists group similar bacteria, but it can make mistakes when bacteria mix genes too much.
Methodology
The reliability of eBURST was evaluated using simulated bacterial populations with known ancestry and varying levels of recombination.
Potential Biases
Potential bias due to oversampling of certain strains in real datasets.
Limitations
The study primarily used simulated data, which may not fully capture the complexities of real bacterial populations.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website