Recognizing the pseudogenes in bacterial genomes

Pseudogenes are now known to be a regular feature of bacterial genomes and are found in particularly high numbers within the genomes of recently emerged bacterial pathogens. As most pseudogenes are recognized by sequence alignments, we use newly available genomic sequences to identify the pseudo-genes in 11 genomes from 4 bacterial genera, each of which contains at least 1 human pathogen. The numbers of pseudogenes range from 27 in Staphylococcus aureus MW2 to 337 in Yersinia pestis CO92 (e.g. 1–8 % of the annotated genes in the gen-ome). Most pseudogenes are formed by small frameshifting indels, but because stop codons are A + T-rich, the two low-G + C Gram-positive taxa (Streptococcus and Staphylococcus) have relatively high fractions of pseudogenes generated by non-sense mutations when compared with more G + C-rich genomes. Over half of the pseudogenes are produced from genes whose original functions were annotated as ‘hypothetical ’ or ‘unknown’; however, several broadly distributed genes involved in nucle-otide processing, repair or replication have become pseudogenes in one of the sequenced Vibrio vulnifi-cus genomes. Although many of our comparisons involved closely related strains with broadly overlap-ping gene inventories, each genome contains a lar-gely unique set of pseudogenes, suggesting that pseudogenes are formed and eliminated relatively rapidly from most bacterial genomes