Transfer bootstrap expectation vs. UFBoot #422
-
Hi everyone, I'd like to calculate transfer bootstrap expectation (TBE) support values for my ML tree. This comment by @bqminh from the old IQ-TREE repo suggests that TBE shouldn't be used in conjunction with ultrafast bootstrap approximation (UFBoot), but the links provided in support of this statement point to a now-defunct Google group and do not resolve anymore. Unfortunately, it seems like the relevant information is not available anywhere else, or at least I have been unable to find it. Could someone please briefly summarize why it is inappropriate to combine these two ways of calculating branch support, and if there are any other relevant recommendations for using TBE within IQ-TREE? |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
Hi there, the TBE (https://www.nature.com/articles/s41586-018-0043-0) was introduced in the context of standard Felsenstein's bootstrap and all the analysis and results published are associated with standard bootstrap. The reason why I'm reluctant to use TBE with UFBoot is because there is no study about the interpretation in this context, which is important because UFBoot values has a different meaning compared with standard bootstrap supports. However, TBE can be already used with -b option (standard bootstrap) in IQ-TREE, simply by adding --tbe option. |
Beta Was this translation helpful? Give feedback.
Hi there, the TBE (https://www.nature.com/articles/s41586-018-0043-0) was introduced in the context of standard Felsenstein's bootstrap and all the analysis and results published are associated with standard bootstrap. The reason why I'm reluctant to use TBE with UFBoot is because there is no study about the interpretation in this context, which is important because UFBoot values has a different meaning compared with standard bootstrap supports.
However, TBE can be already used with -b option (standard bootstrap) in IQ-TREE, simply by adding --tbe option.