Table 2.

Supertree topology estimation error on simulated datasets, measured using the Robinson-Foulds error rate, expressed as a percentage

Method	100	100	100	100	500	500	500	500	1000	1000	1000	1000
Scaffold %	20	50	75	100	20	50	75	100	20	50	75	100
# Replicates	9	10	10	10	8	10	10	10	10	10	10	10
ASTRAL	11.7	14.0	11.6	10.0	15.3	14.8	12.7	11.2	16.9	15.7	13.6	11.6
ASTRAL-enhanced	11.8	13:1	11.5	10.0	14.8	14.1	12.6	11.2	$\mathbf{16.3}$	$\mathbf{15.1}$	13.5	11.6
ASTRID	15.8	18.7	17.1	9.6	26.0	50.1	45.4	$\mathbf{10.5}$	35.6	58.1	52.0	$\mathbf{11.2}$
MRL	13.6	13.6	11.2	10.8	15.4	14.3	12.1	11.2	17.4	$\mathbf{15.1}$	13.5	12.2
MulRF	22.1	26.0	15.3	9.3	46.9	40.3	27.4	12.6	–	–	–	–
PluMiST	25.9	16.6	11.5	9.3	35.4	29.5	22.4	10.9	–	–	–	–
FastRFS-basic	13.5	14.3	$\mathbf{10.5}$	$\mathbf{9.1}$	14.5	14.3	12.4	11.1	17.3	15.6	13.5	12.0
FastRFS-enhanced	13.5	13.4	10.6	9.3	$\mathbf{14.3}$	$\mathbf{13.9}$	$\mathbf{12.0}$	10.8	16.7	$\mathbf{15.1}$	$\mathbf{13.4}$	11.8

The best result for each model condition is boldfaced. No results are shown for PluMiST or MulRF on the 1000-taxon simulated datasets due to running time limitations for these methods. Results are averaged over the completed replicates.