| Scaffold Density |
---|
Method | 20 | 50 | 75 | 100 | Average |
---|
MRP(PAUP*) | 20.7 (1.1) | 17.7 (0.8) | 16.1 (0.8) | 11.7 (0.9) | 16.2 (0.7) |
MRP(TNT) | 27.9 (1.6) | 19.1 (1.3) | 14.9 (1.1) | 11.7 (0.8) | 17.6 (1.1) |
MRL(RAxML) | 15.7 (0.7) | 14.1 (0.6) | 13.8 (1.0) | 11.9 (0.8) | 13.7 (0.5) |
SCM |
5.9(0.5)
|
5.4(0.4)
|
4.9(0.6)
|
4.4(0.6)
|
5.1(0.3)
|
SuperFine+MRP(PAUP*) | 14.4 (0.6) | 13.2 (0.6) | 12.7 (0.8) | 11.6 (0.8) | 12.8 (0.4) |
SuperFine+MRP(TNT) | 14.4 (0.7) | 13.0 (0.6) | 12.6 (0.8) | 11.8 (0.8) | 12.8 (0.4) |
SuperFine+MRL(RAxML) | 14.8 (0.7) | 13.5 (0.5) | 12.9 (0.8) | 11.9 (0.8) | 13.1 (0.4) |
- We show average false positive rates (reported as %) on the 1000-taxon datasets. False positive rate is calculated as total number of FP edges in the estimated tree divided by the total number of internal edges in the internal tree. Each simulated dataset has 25 clade-based source trees and 1 scaffold tree. The scaffold density is the percentage of the full taxon set that is present in the scaffold tree. The standard error is shown in parenthesis. n = 7 for the 20% scaffold density, n = 9 for the 50% scaffold density, and n = 10 for the remaining scaffold densities. n = 36 for the average. The lowest false positive rate for each scaffold density is shown in bold.