Assigning Escherichia coli strains to phylogenetic groups: multi‐locus sequence typing versus the PCR triplex method

Summary It is well recognized that Escherichia coli consists of a number of distinct phylo‐groups and that strains of the different phylo‐groups vary in their ecological niches, life‐history characteristics and propensity to cause disease. Consequently, much can be learnt by assigning a strain of E. coli to one of the recognized phylo‐groups. A triplex PCR‐based method that enables strains of E. coli to be assigned to a phylo‐group using a dichotomous key approach based on the presence or absence of two genes ( chuA and yjaA ) and an anonymous DNA fragment (TSPE4.C2) has been developed. However, the accuracy with which this method assigns strains to their correct phylo‐group has not been adequately evaluated. Consequently, 662 strains of E. coli were characterized using a multi‐locus sequence typing approach. Unsupervised population assignment algorithms were used to assign strains to phylo‐groups based on the multi‐locus sequence typing data. The analyses revealed that 85–90% of E. coli strains can be assigned to a phylo‐group and that 80–85% of the phylo‐group memberships assigned using the Clermont method are correct. However, the accuracy with which strains are assigned to the correct phylo‐group depends on their Clermont genotype. For example, strains yielding a Clermont genotype consistent with phylo‐groups B1 and B2 are assigned correctly 95% of the time. Strains failing to yield any PCR products using the Clermont method are seldom members of phylo‐group A and strains with such a genotype should not be assigned to a phylo‐group.