Additional file 1 of A large-scale genomically predicted protein mass database enables rapid and broad-spectrum identification of bacterial and archaeal isolates by mass spectrometry

Additional file 1: Supplementary Material 1. The set of Pfam models used for extracting coding proteins likely relevant to ribosomal proteins. Figure S1. Relationship between average nucleotide identity (ANI) and genome-wide peak matching (PM) score. Figure S2. Relationship between taxonomic distance and genome-wide peak matching (PM) score. Figure S3. Relationship between taxonomic distance and genome-wide peak matching (PM) score for three genome assemblies with among the highest number of genes, or predicted mass peaks, in the GPMsDB (Kosakonia sp002886105: 8,428 predicted mass peaks with m/z values of 2,000–20,000; Streptomyces violaceus: 8,375; Actinoplanes liguriensi: 8,171). Figure S4. Effect of scoring scheme on identification accuracy as evaluated with simulated mass peak lists. Figure S5. Accuracy of identification using scoring scheme I for simulated peak lists with varying degrees of “noise”. Figure S6. Accuracy of identification using scoring scheme III for simulated peak lists containing varying numbers of ribosomal protein peaks. Figure S7. Schematic of the GPMsDB-tk. Figure S8. Identification results for MALDI-TOF MS peak lists of 71 bacterial reference strains. Figure S9. Identification results for MALDI-TOF MS peak lists of 10 archaeal strains. Figure S10. Identification results for MALDI-TOF MS peak lists of 13 strains of Actinomycetota. Figure S11. Identification results for MALDI-TOF MS peak lists for 74 different Acinetobacter strains. Figure S12. Same as for Figure S11, except that all genomes in the GPMsDB were searched by specifying command line option -a as “all”. Figure S13. Identification results for MALDI-TOF MS peak lists for 24 different strains of Cutibacterium acnes