Rethinking the applicability of Tenover criteria: a model algorithm and a new dendrogram approach for the direct and wider application
Abstract number: 1733_1350
Acuner I.C., Eroglu C.
Objectives: Modified Tenover criteria (MTC) propose the difference that is greater than two genetic events (i.e. >eight band difference) as the criterion for unrelated isolates. However, in the investigation of outbreaks, PFGE patterns of the isolates are mostly evaluated on the basis of various similarity indexes to construct similarity matrices, and then to generate dendrograms through various methods. One of the most used dendrogram generation methods, UPGMA, together with the assured statistical significance by bootstrapping, is claimed to produce results in concordance with MTC. Although their biological basis is more relevant than the other methods, Tenover criteria didn't find much ground for direct application. In this study, direct applicability of MTC by the use of an algorithm to detect the related isolates within the limit of one genetic event difference has been investigated.
Methods: Based on the variables such as genetic variation mechanism with or without affecting the restriction site existence or emergence, relative event position and length, genetic event number, band difference, position, and intensity, in a set of hypothetical PFGE genotypes and in the Escherichia coli strains with known whole genome sequences, random and calculated PFGE patterns are generated, respectively. Consequently, related genotypes up to three genetic event differences have been investigated by the application of an algorithm that compares each PFGE genotype, to the ``index'' PFGE genotype and a new dendrogram approach has been adapted.
Results: One genetic event differences that may occur with different mechanisms as described by MTC were succesfully detected by the developed algorithm and shown to be distinguishable from each other. MTC display a special pairwise symmetry in the genetic event types. MTC cause triangle inequality and hidden histories like the other above mentioned methods. The selection of the index isolate of an epidemic seems critical. MTC can not exclude the probability of the occurrence of PFGE genotypes that may have greater than two genetic event differences with lower than nine band differences. Direct application of MTC in two and more genetic events may become so complex that it does not allow clear-cut interpretation.
Conclusion: It has been shown that MTC can be applicable for automated direct use in the analysis of large sets of epidemic isolates by PFGE through an algorithm within the limit of one genetic event difference.
|Session name:||European Society of Clinical Microbiology and Infectious Diseases|
|Location:||ICC, Munich, Germany|
|Back to top|