An iterative and automated computational pipeline for untargeted strain-level identification using MS/MS spectra from pathogenic samples

Mathias Kuhring; Joerg Doellinger; Andreas Nitsche; Thilo Muth; Bernhard Y. Renard

The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/812313 doi: bioRxiv preprint Since Pipasic is sensitive to a high number of taxa, we limit the input to taxa with a minimum of two hits as well as to the 100 most abundant taxa. Expressed proteins per taxa are extracted according to the taxonomic classification and digested peptides with a minimum length of six amino acids are prepared using trypsin digestion 58 . PSMs and tryptic peptides are then passed to Pipasic to obtain corrected relative counts. We perform all three strategies on several viral and bacterial MS/MS spectra samples with available strainlevel knowledge. This includes a Cowpox virus (Brighton Red) strain (acquired in-house), an Avian infectious bronchitis virus (strain Beaudette CK) 59 and a Bacillus subtilis BSN238 60 that we will refer to as cowpox, bronchitis and bacillus sample, respectively. B. subtilis BSN238 is a transgenic organism resulting from horizontal gene transfer (HGT) of the DivIVA protein from Listeria monocytogenes strain EGD-e to Bacillus subtilis subsp. subtilis str. 168. Since B. subtilis BSN238 is neither yet present in the NCBI Taxonomy nor in the NCBI Protein database and only one protein is modified, we expect B. subtilis subsp. subtilis str. 168 to be selected as final strain candidate. Bacillus samples are examined twice, once completely with 28902 spectra (bacillus all) and once randomly reduced to 1000 spectra (bacillus 1k) to improve performance with respect to the vast bacterial search space. A detailed description of the cowpox sample acquisition and the search parameters for all samples is provided in Supplementary item 1.