Identifying Enterobacter sakazakii Using Artificial Neural Networks
Author Information
Author(s): Carol Iversen, Lee Lancashire, Michael Waddington, Stephen Forsythe, Graham Ball
Primary Institution: The Nottingham Trent University
Hypothesis
Can Artificial Neural Networks improve the identification of Enterobacter sakazakii from closely related species?
Conclusion
The study demonstrates that Artificial Neural Networks can significantly enhance the identification accuracy of Enterobacter sakazakii, providing a foundation for rapid diagnostic tests.
Supporting Evidence
- Artificial Neural Networks achieved 99.3% correct discrimination between E. sakazakii and closely related species.
- The models showed 100% specificity for phenotypic data.
- Three key regions of the rDNA sequence were identified as important for species discrimination.
Takeaway
Scientists used computer programs to help identify a germ that can make babies sick, and they found a really good way to tell it apart from similar germs.
Methodology
Artificial Neural Networks were applied to biochemical and 16S rDNA data from 282 strains of Enterobacteriaceae, including 189 E. sakazakii isolates.
Potential Biases
Potential misclassification of strains due to overlapping characteristics among species.
Limitations
The study may not account for all possible strains of Enterobacteriaceae and their biochemical characteristics.
Participant Demographics
Strains were from diverse food, clinical, and environmental sources worldwide.
Statistical Information
P-Value
0.0001
Confidence Interval
95%
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website