Predicting Genomic Islands in Microbial Genomes
Author Information
Author(s): Stephan Waack, Oliver Keller, Roman Asper, Thomas Brodag, Carsten Damm, Wolfgang Florian Fricke, Katharina Surovcik, Peter Meinicke, Rainer Merkl
Primary Institution: Institut für Informatik, Universität Göttingen
Hypothesis
Can hidden Markov models effectively predict genomic islands in prokaryotic genomes based on codon usage?
Conclusion
The SIGI-HMM tool is effective for identifying genomic islands in microbial genomes and predicting the origins of horizontally transferred genes.
Supporting Evidence
- SIGI-HMM was validated against known genomic islands and showed consistent predictions.
- The tool is publicly available and can analyze any genome formatted in EMBL.
- Benchmark tests confirmed the accuracy of SIGI-HMM's predictions.
Takeaway
Scientists created a computer program that helps find special pieces of DNA in bacteria that come from other organisms, which can help them adapt to new environments.
Methodology
The program SIGI-HMM analyzes codon usage of genes in a genome to predict genomic islands and their potential donors using a hidden Markov model.
Potential Biases
The algorithm may misclassify highly expressed genes as horizontally transferred due to their distinct codon usage.
Limitations
The method may not identify genomic islands with codon usage similar to the host's, and it relies solely on codon usage analysis.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website