HIPSTR: highest independent posterior subtree reconstruction in TreeAnnotator X.

Area of Science:

• Phylogenetics
• Computational Biology
• Evolutionary Biology

Background:

• Bayesian phylogenetic and phylodynamic studies commonly use summary phylogenies to represent posterior distributions of trees.
• The maximum clade credibility (MCC) tree is a frequently employed method for summarizing phylogenetic information.

Purpose of the Study:

• To introduce and evaluate a novel summary tree method, the highest independent posterior subtree reconstruction (HIPSTR).
• To compare the performance of HIPSTR against the traditional MCC tree method in terms of clade support and computational efficiency.

Main Methods:

• Developed and implemented HIPSTR and its extension MrHIPSTR within the TreeAnnotator X software.
• Applied HIPSTR and MCC methods to reconstruct summary trees from Ebola virus and SARS-CoV-2 datasets.
• Evaluated clade support and computational performance of HIPSTR and MCC.

Main Results:

• HIPSTR consistently yields summary trees with higher supported clades compared to MCC trees across multiple viral datasets.
• MCC trees frequently omit clades with high posterior probability (≥0.95) and moderate to high support (≥50%), while HIPSTR and MrHIPSTR demonstrate near-perfect performance.
• HIPSTR and MrHIPSTR show favorable computational performance over MCC within TreeAnnotator X.
• Initial comparisons with the CCD0-MAP algorithm suggest mixed results requiring further investigation.

Conclusions:

• HIPSTR is a superior method for reconstructing time-calibrated summary phylogenies in Bayesian analyses, offering enhanced clade support over MCC.
• The updated TreeAnnotator X with HIPSTR and MrHIPSTR provides faster computational routines for summary tree estimation.
• HIPSTR represents a significant advancement in summarizing complex phylogenetic information, particularly for viral evolution studies.