Table 7. Q and TC scores on SMART.
| Method | Q | TC | Significance |
|---|---|---|---|
| MUSCLE | 0.855 | 0.537 | 0.07 |
| NWNSI | 0.848 | 0.546 | 0.03 |
| MUSCLE-p | 0.836 | 0.505 | 0.54 |
| T-Coffee | 0.835 | 0.503 | 0.16 |
| CLUSTALW | 0.823 | 0.504 | 0.07 |
| FFTNS1 | 0.817 |
The average Q and TC accuracy scores over the 267 reference alignments in SMART that have no more than 100 sequences are given. The last column is the P-value of the difference between the method in a row and the method in the next row, measured on the Q score. The P-value for MUSCLE > T-Coffee is 0.0004 on Q and 0.01 on TC; the P-value for NSI > T-Coffee is 0.19 on Q and 0.0002 on TC. The difference between MUSCLE and NWNSI is only weakly significant on the Q score (P = 0.07) and is not significant on the TC score (P = 0.3).