Analyzing the Prevalence of Newly Discovered Bacterial Defense Genes and Approaches to their Quantification

Abstract

Bacteria possess a variety of defense systems to protect themselves from bacteriophage infection. These systems have been exploited to create widely used tools in both experimental and therapeutic contexts. Defense system genes tend to be clustered on the genome within ‘defensive islands.’ By searching for genes overrepresented on those islands, an enormous variety of defense systems were recently discovered. The Prokaryotic Antiviral Defense System (PADS) Arsenal is the first database to incorporate these newly discovered systems into a comprehensive list of all bacterial defense genes. This thesis analyzes the defense genes identified in the PADS Arsenal and the efficacy of the PSI-BLAST quantification tool for future defense gene identification. Analysis of the PADS Arsenal in 62,164 bacteria uncovered a high prevalence of defense system genes and a tendency for these genes to co-occur in the same genome, which is partially dependent on genome length. Despite their high prevalence in general, this analysis also identified a small number of bacterial genomes with no known defensive genes. The PSI- BLAST search tool demonstrated low recall and precision in identifying “cyclase,” a defense gene from the novel cyclic-oligonucleotide-based anti-phage signaling system. Transitioning to matrix-based search methods and focusing identifications of defense mechanisms on complete systems will likely increase the recall and precision of future defense system quantifications. Our findings provide a comprehensive view of the distribution of anti-phage defense systems in 62,164 bacterial genomes and evaluates currently used methods for their identification and quantification.