Skip to main content
. 2023 Jul 19;18:6. doi: 10.1186/s13015-023-00230-6

Fig. 1.

Fig. 1

Comparison of the choice of branch support and the choice of branch length normalization strategy for wASTRID-s and wASTRID-pl respectively on the training data, showing the species tree topological error rates (nRF error). “SH-like” is FastTree SH-like support. “aBayes” is IQ-TREE approximate Bayesian support. “BS” is bootstrap support using 100 FastTree trees. The x-axis varies in the number of genes k in the input. Results are shown averaged across ten replicates. Error bars show standard error