Speaker
Description
Core genome multilocus sequence typing (cgMLST) is a widely applied method for bacterial pathogen surveillance and outbreak investigations, in which pairwise genetic distances are commonly used to calculate relatedness between samples through single-linkage clustering. However, cluster definitions heavily depend on the underlying typing scheme and allele calling strategy.
Using datasets of Mycobacterium tuberculosis and Listeria monocytogenes, we evaluated the impact of open reading frame (ORF) filtering on cgMLST-based clustering. We directly compared clustering results calculated from complete and ORF-filtered cgMLST profiles generated by multiple established allele calling tools. For this, an element-centric similarity score was utilized.
We determined allelic distance thresholds required to obtain comparable clustering results across different allele calling methods. In summary, ORF-filtering reduced pairwise allelic distances between isolates by increasing cgMLST missingness. Consequently, complete profiles produced clusters with potentially increased genomic resolution compared to ORF-filtered profiles at fixed distance thresholds.
We observed that ORF-filtering substantially affects cgMLST-based distances and outbreak cluster definitions. Re-evaluating clustering thresholds based on each method enables a more consistent, meaningful intersectoral exchange regarding outbreak analyses and routine genomic surveillance.
Keywords
Bioinformatics, cgMLST, Clustering, Outbreak investigations
| Registration ID | OHS26-70 |
|---|---|
| Professional Status of the Speaker | Graduate Student |
| Junior Scientist Status | Yes, I am a Junior Scientist. |
Author
Co-authors
External references
- 139
