Systematic derivation of marker sets for staphylococcal cassette chromosome mec typing

Abstract

The aim of this study was to identify optimized sets of genotyping targets for the staphylococcal cassette chromosome mec (SCCmec). We analyzed the gene contents of 46 SCCmec variants in order to identify minimal subsets of targets that provide useful resolution. This was achieved by firstly identifying and characterizing each available SCCmec element based on the presence or absence of 34 binary targets. This information was used as input for the software "Minimum SNPs," which identifies the minimum number of targets required to differentiate a set of genotypes up to a predefined Simpson's index of diversity (D) value. It was determined that 22 of the 34 targets were required to genotype the 46 SCCmec variants to a D of 1. The first 6, 9, 12, and 15 targets were found to define 21, 29, 35, and 39 SCCmec variants, respectively. The genotypes defined by these marker subsets were largely consistent with the relationships between SCCmec variants and the accepted nomenclature. Consistency was made virtually complete by forcing the computer program to include ccr1 and ccr5 in the target set. An alternative target set biased towards discriminating abundant SCCmec variants was derived by analyzing an input file in which common SCCmec variants were repeated, thus ensuring that markers that discriminate abundant variants had a large effect on D. Finally, it was determined that mecA single nucleotide polymorphisms (SNPs) can increase the overall genotyping resolution, as different mecA alleles were found in otherwise identical SCCmec variants.

FIG. 1.

Illustration of the strategy used to identify sets of resolution-optimized binary markers. Each genotype of interest is characterized based on the presence (T) or absence (A) of the total set of binary markers (five are used in this example). Thus the binary gene configuration is converted into a pseudo-DNA sequence composed of A's and T's. The alignment of pseudo-DNA sequences is then analyzed using Minimum SNPs for combinations of binary markers that maximize the Simpson's index of diversity (D) (33). In the example shown above, binary genes 2 and 4 provide a D of 1, i.e., they completely resolve the three genotypes. In the present study, 46 SCCmec types were defined by the presence or absence of 34 binary markers. Even with this relatively small data set, manual identification of resolution-optimized marker sets is extremely difficult.