Utilizing large and diverse bacterial genome datasets to improve the detection and identification of Streptococcus pneumoniae via PCR-based diagnostics.

The accurate identification of Streptococcus pneumoniae (pneumococcus) is crucial for diagnostics and surveillance but is complicated by the use of molecular assays that may also detect non-pneumococcal Streptococcus (NPS) species. Therefore, the aim of this study was to use a combination of in silico and in vitro analyses to evaluate PCR assays for the molecular detection and identification of pneumococci. A diverse dataset of over 9,300 pneumococcal and NPS genomes was investigated in silico to determine the sensitivity and specificity of assays for seven recommended gene targets: lytA, piaB, ply, psaA, Spn9802, SP2020 and Xisco. These in silico findings were used to design new diagnostic assays for two targets, Xisco and SP2020. The new assays were evaluated in vitro using three sets of isolates, one of which was selected based upon evidence for sequence diversity from a second in silico investigation of over 6,000 pneumococcal genomes sequenced by the United Kingdom Health Security Agency. Experimentally, the new Xisco and SP2020 assays were compared to published assays for lytA and piaB. The in vitro specificity was 100% (95% CI, 98.7-100%) across all assays. The in vitro sensitivity was 100% (95% CI, 98.5-100%) for lytA, SP2020_new and the Xisco assays and 99.6% (95% CI, 97.8-100%) for piaB. The new assays were found to be highly sensitive and specific and able to detect as few as two pneumococcal genome copies per quantitative PCR reaction. Overall, this study demonstrated the value of performing large-scale in silico genomic analyses of diagnostic targets, followed by in vitro testing that was specifically designed to account for global pneumococcal population-level diversity.