Jackknife and Bootstrap Tests of the Composition Vector Trees

Zuo, Guanghong; Xu, Zhao; Yu, Hongjie; Hao, Bailin

This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). However, several anonymous referees of our previous publications and a few users of the CVTree web server (2, 3) have repeatedly raised the questions of statistical support of the CVTrees. Correspondingly, we have shown a few bootstrap results in Qi et al (1) and Wang et al (9) without giving details. In these papers, we described the method of our bootstrap, jackknife and anti-jackknife (see below) tests of CVTrees. We showed the results on the examples of virus, bacteria and fungi. Since full-fledged bootstrap/jackknife tests are extremely time-consuming and soon come the epoch of inferring phylogeny for thousands of species in not-too-distant future, our ultimate goal is to advocate the viewpoint that one may give up routine statistical re-sampling tests for the robust and stable CV method, and rely on direct comparison of phylogeny with taxonomy to justify the inferred trees.