Table 1.
Paired tests for detection quality on individual queries
| PSI-BLAST | COMPASS | HHsearch no SS | prof_ss | HHsearch | |
| PSI-BLAST | |||||
| 0.09 | -0.00 | -6.39e-317 | -0.00 | -0.00 | |
| 0.18 | -0.00 | -0.00 | -1.03e-317 | -0.00 | |
| 0.31 | -2.85e-179 | -2.71e-248 | -2.19e-180 | -1.34e-276 | |
| 0.12 | -0.00 | -0.00 | -0.00 | -0.00 | |
| COMPASS | |||||
| 0.13 | +0.00 | -2.96e-41 | -8.63e-200 | -0.00 | |
| 0.22 | +0.00 | -3.48e-151 | -8.48e-48 | -0.00 | |
| 0.38 | +2.85e-179 | +4.10e-07 | +3.89e-22 | -1.18e-52 | |
| 0.24 | +0.00 | +4.75e-141 | -0.00 | -0.00 | |
| HHsearch no SS | |||||
| 0.14 | +6.39e-317 | +2.96e-41 | +0.00 | -0.00 | |
| 0.27 | +0.00 | +3.48e-151 | +6.92e-10 | -2.60e-274 | |
| 0.38 | +2.71e-248 | -4.10e-07 | +4.44e-22 | -5.06e-139 | |
| 0.22 | +0.00 | -4.75e-141 | -0.00 | -0.00 | |
| prof_ss | |||||
| 0.14 | +0.00 | +8.63e-200 | -0.00 | -0.00 | |
| 0.21 | +1.03e-317 | +8.48e-48 | -6.92e-10 | -2.77e-258 | |
| 0.37 | +2.19e-180 | -3.89e-22 | -4.44e-22 | -1.76e-15 | |
| 0.45 | +0.00 | +0.00 | +0.00 | +0.00 | |
| HHsearch | |||||
| 0.19 | +0.00 | +0.00 | +0.00 | +0.00 | |
| 0.28 | +0.00 | +0.00 | +2.60e-274 | +2.77e-258 | |
| 0.38 | +1.34e-276 | +1.18e-52 | +5.06e-139 | +1.76e-15 | |
| 0.34 | +0.00 | +0.00 | +0.00 | -0.00 |
Methods are compared by sensitivity values at 50% selectivity, calculated separately for each query. For each pair of methods, P-values of paired Wilcoxon rank test are shown for four types of criteria for true/false positive distinction (from top to bottom): reference-dependent without alignment quality and with alignment quality, and reference-independent global and local. Plus and minus signs by the P-values denote respectively positive and negative difference for the method indicated on the left Vs the method indicated on the top. The left column includes mean values over all queries for each method.