• There are high differences in the numbers of annotated genes, which don't correspond to the real genome similarity among bacterial strains (see Figure 1).
• Estimations of the core-genome size based on only original annotations are underestimeted (see Figure 2).

Figure 1:This plot presents numbers of annotated genes and numbers of the multigenes after the closure procedure applied to E. coli strains. On the x-axis strains are listed (from left to right) in descending order of their genome length. The blue line and the red line present respectively the number of annotated genes and the number of multigenes (after the closure) for each strain. The green line presents the number of multigenes after the closure and after applied post-processing removal of multigenes shorter than 200 nucleotides length.

Figure 2: Core vs. pangenome plots of 41 E. coli strains calculated using original annotations and multigene anno- tations, predicted by CAMBer. Strains are sorted (from left to right) in descending order of their genome sizes. Violet and green (coregenome-annot and pangenome-annot) lines connect cumulative numbers of core and pangenome sizes using annotated genes, while red and blue (coregenome-multi and pangenome-multi) lines connect cumulative numbers of core and pangenome sizes using multigenes after the closure procedure. The proportion of core genome to pangenome size has risen from 18% to 25% after the closure.

CAMBer results for the E. coli case study

• Excel table with CAMBer results before the refinement procedure.
• Excel table with CAMBer results after the refinement procedure.

CAMBer results for the S. aureus case study

• Excel table with CAMBer results before the refinement procedure.
• Excel table with CAMBer results after the refinement procedure.

CAMBer results for the M. tuberculosis case study

• Excel table with CAMBer results before the refinement procedure.
• Excel table with CAMBer results after the refinement procedure.