Pseudogenes within each genome were identified by comparing its genome contents with that of each of the other sequenced strains within the particular taxonomic group. For each genome, we first retrieved the set of the annotated proteins from GenBank, which were used to query the nucleotide sequence of the other genomes within the particular group using TBLASTN (23 (link)). For example, the proteins of Y.pseudotuberculosis were used to query the genome sequences of Y.pestis CO92, Y.pestis KIM and Y.pestis 91001. We then applied the program Ψ–Φ (5 (link)) on the BLAST outputs to recover candidate pseudogenes in each genome. This program allows the specification of any BLAST score and % identity cut-offs, and for our comparisons, proteins from two genomes were considered to be homologous if their BLAST score reached an E-value <10−15 and their level of protein identity was >79%. In the case of Vibrio, the strains/species examined were not as closely related, so we applied different thresholds (E-values <10−10 and a minimal percentage of protein identity of 49%) in order to identify homologous sequences. This program retrieves pseudogenes that result from nonsense mutations, frameshifts generated by small insertions or deletions, large insertions (such as those resulting from transposable elements) and truncations of any specified length as well as any incorrectly annotated spacers that resulted from degradation of a gene. Lists of candidate pseudogenes were curated manually, and the disrupting mutations were determined by aligning the nucleotide sequences of putative pseudogenes with their functional counterparts using CLUSTALW 1.8 (24 (link)).