Table 3.
The number of incongruent topologies detected by the Shimodaira-Hasegawa (SH) test.
| Bacillaceae | |||||
| core | non-core | ||||
| N | rejected | total | rejected | total | p-value |
| 4 | 3 (3.5) | 85 | 32 (14.0) | 228 | 0.004573 |
| 5 | 11 (6.2) | 176 | 38 (21.6) | 176 | 2.170e-05 |
| 6 | 17 (5.4) | 314 | 34 (29.6) | 115 | 1.899e-10 |
| 7 | 32 (4.3) | 739 | 9 (15.5) | 58 | 0.001705 |
| total | 63 (4.8) | 1314 | 113 (19.6) | 577 | <2.2e-16 |
| Enterobacteriaceae | |||||
| core | non-core | ||||
| N | rejected | total | rejected | total | p-value |
| 4 | 6 (5.0) | 121 | 19 (17.1) | 111 | 0.002496 |
| 5 | 12 (5.1) | 235 | 21 (24.4) | 86 | 2.784e-06 |
| 6 | 28 (5.7) | 490 | 9 (34.6) | 26 | 2.431e-05 |
| 7 | 45 (4.1) | 1099 | 0 (0.0) | 8 | 1 |
| total | 91 (4.7) | 1945 | 49 (21.2) | 231 | 6.409e-16 |
The SH test was performed for all OGs (both core and non-core) that have a one-to-one correspondence using the TREE-PUZZLE program with the ML topology (or its subtree) of the concatenated core sequences (Figure 8A) as a reference tree, and the numbers of rejected (significantly incongruent) cases were compared between core and non-core classes using the Fisher's exact test. The first column (N) indicates effective number of genomes where genomes of closely related species were counted only once. The column "rejected" indicates the number of cases rejected by the SH test at a significance level of 0.05; the number in parentheses is the percentage of rejected cases. The last column indicates the p-value of a one-sided Fisher's exact test.