s. 0/4 431 176 n.s. 0/4 Rhizobium leguminosarum 2 3678 4063 n.s. 2/4 148 176 n.s. 2/4 Rickettsia bellii 2 1277 850 ** 0/25 219 1 ** 0/25 Rickettsia rickettsii 2 1221 850 ** 0/25 93 1 ** 0/25 Shigella boydii 2 3170 2989 ** 1/17 95 12 ** 0/17 Shigella flexneri 3 3255 2770 ** 0/25 130 6 ** 0/25 Staphylococcus aureus 14 1917 1486 ** 0/25 157 0 ** 0/25 Staphylococcus epidermidis 2 2080 1798 ** 0/25 131 0 ** 0/25 Streptococcus agalactiae 3 1688 1019 ** 0/25 156 0 – 0/25 Streptococcus pneumoniae 6 1543 922 ** 0/25 150 0 -

0/25 Streptococcus pyogenes 13 1348 811 ** 0/25 49 0 – 0/25 Streptococcus suis #Avapritinib concentration randurls[1|1|,|CHEM1|]# 2 1971 1087 ** 0/25 336 0 ** 0/25 Streptococcus thermophilus 3 1359 1019 ** 0/25 145 0 – 0/25 Vibrio cholerae 2 3384 2764 ** 1/25 Selleck AZD5582 425 20 ** 0/25 Vibrio fischeri 2 3380 2764

** 1/25 447 20 ** 0/25 Vibrio vulnificus 2 3882 2764 ** 0/25 321 20 ** 0/25 Xanthomonas campestris 4 3376 2818 ** 0/25 49 4 ** 0/25 Xanthomonas oryzae 3 3276 2915 ** 5/25 299 0 ** 0/25 Yersinia pestis 7 2986 2717 ** 4/25 21 0 ** 0/25 Yersinia pseudotuberculosis 4 3424 3003 ** 0/25 21 0 ** 0/25 For the meanings of each column, see Table 3. The primary purpose of this section was to investigate the utility of this cohesiveness analysis for identifying bacterial species that might be misclassified. A cursory reading of Tables 3 and 4 revealed that, while most species satisfied both of the above criteria, some species either had core or unique proteomes that were not significantly larger than the average of the random groups, or had several corresponding random groups that had larger core or unique proteomes than the species itself. A lack of cohesiveness in the proteomes of a given species indicates that its taxonomic classification may need revisiting. However, these results must be interpreted with caution. A closer look at these species revealed that the classification Glycogen branching enzyme of some really

did appear to warrant re-examination, whereas the apparent lack of cohesiveness of others had alternative explanations. In the following paragraphs, we discuss several examples. First, we describe the cohesiveness results for Bacillus anthracis, which is indeed proteomically cohesive based on Tables 3 and 4. Next, we discuss Rhizobium leguminosarum and Yersinia pestis, both of which look uncohesive based on these tables but whose lack of cohesiveness can readily be explained. Finally, we look at two species that probably do warrant reclassification, Bacillus cereus and Bacillus thuringiensis. As an example of reading Tables 3 and 4, consider the first row of Table 3, which contains B. anthracis. The core proteome of the three sequenced B. anthracis isolates contained 4941 proteins.