Development of a Web Tool for Escherichia coli Subtyping Based on <i>fimH</i> Alleles

The aim of this study was to construct a valid publicly available method for <i>in silico fimH</i> subtyping of <i>Escherichia coli</i> particularly suitable for differentiation of fine-resolution subgroups within clonal groups defined by standard multilocus sequence typing (MLST). FimTyper was constructed as a FASTA database containing all currently known <i>fimH</i> alleles. The software source code is publicly available at https://bitbucket.org/genomicepidemiology/fimtyper, the database is freely available at https://bitbucket.org/genomicepidemiology/fimtyper_db, and a service implementing the software is available at https://cge.cbs.dtu.dk/services/FimTyper FimTyper was validated on three data sets: one containing Sanger sequences of <i>fimH</i> alleles of 42 <i>E. coli</i> isolates generated prior to the current study (data set 1), one containing whole-genome sequence (WGS) data of 243 third-generation-cephalosporin-resistant <i>E. coli</i> isolates (data set 2), and one containing a randomly chosen subset of 40 <i>E. coli</i> isolates from data set 2 that were subjected to conventional <i>fimH</i> subtyping (data set 3). The combination of the three data sets enabled an evaluation and comparison of FimTyper on both Sanger sequences and WGS data. FimTyper correctly predicted all 42 <i>fimH</i> subtypes from the Sanger sequences from data set 1 and successfully analyzed all 243 draft genomes from data set 2. FimTyper subtyping of the Sanger sequences and WGS data from data set 3 were in complete agreement. Additionally, <i>fimH</i> subtyping was evaluated on a phylogenetic network of 122 sequence type 131 (ST131) <i>E. coli</i> isolates. There was perfect concordance between the typology and <i>fimH</i>-based subclones within ST131, with accurate identification of the pandemic multidrug-resistant clonal subgroup ST131-<i>H</i>30. FimTyper provides a standardized tool, as a rapid alternative to conventional <i>fimH</i> subtyping, highly suitable for surveillance and outbreak detection.