Genome-based Brucella melitensis Genotyping

Genome assemblies of Italian B. melitensis were genotyped using cgMLST. We used cgMLST because the method generates data that are relatively quick and simple to analyse, and the results can be readily standardized and reported. The cgMLST profiles were assigned using the B. melitensis task template with 2704 target core genes in Ridom SeqSphere+ software, version 4.1.1 (Ridom), as previously described [37 (link)]⁠. Only genomes containing ≥98 % of good targets were accepted for the subsequent analyses. To be accepted as good targets, identified genes need to fulfil the software’s default Target QC parameters, i.e. the same length as the reference gene ± 3 nucleotide triplets, no ambiguities and no frameshifts. A MST was constructed by pairwise comparison of cgMLST alleles. Based on previous data, the cut-off of six allele differences was applied to identify clusters of possibly related genomes [37 (link)]. Default parameters were used and the missing values were excluded in the calculation of distance between genotypes. The public genome dataset and the Italian strains sequenced in this study were genotyped with the Brucella nine locus MLST (MLST-9) scheme available at https://pubmlst.org/brucella/ [42 (link)]⁠ and accessed through Ridom SeqSphere+.