Table 2.
Comparison of tree statistics among character setsa
| A | B | A × B | |
|---|---|---|---|
| Character set | Number of characters | Ave. Δ/character | Tree length |
| nt123 | 6,633(= 1.0) | 35.17 | 233,312 (= 1.00) |
| nt12 | 4,422(0.67) | 8.76 | 38,751 (0.17) |
| nt3 | 2,211(0.33) | 99.79 | 220,645 (0.95) |
| nt2 | 2,211(0.33) | 5.63 | 12,452 (0.05) |
| nt1 | 2,221(0.33) | 15.64 | 34,579 (0.15) |
| noLRall2 | 1,631(0.25) | 6.57 | 10,723 (0.05) |
| LRall2 | 580(0.09) | 26.67 | 15,469 (0.07) |
| noLRall2 + nt2 | 3,842(0.58) | 5.54 | 21,283 (0.09) |
a Based on likelihood analyses of 5-gene, 123-taxon character sets and subsets. The single constraint tree on which parameters are optimized is derived from analysis of nt123.
A, number of characters; B, average number of substitutions per character across all branches; A × B, Tree length = number of substitutions for all characters across all branches. Numbers in parentheses represent comparisons to the nt123 values, defined as 1.0.