Species tree inference from gene trees is an important part of biological research. One confounding factor in estimating species trees is gene duplication and loss, which can lead to gene family trees with multiple copies of the same species. In recent years there have been several new methods developed to address this problem that have substantially improved on earlier methods; however, the best performing methods (ASTRAL-Pro, ASTRID-multi, and FastMulRFS) have not yet been directly compared. In this study, we compare ASTRAL-Pro, ASTRID-multi, and FastMulRFS under a wide variety of conditions. Our study shows that while all three have nearly the same accuracy under most conditions, ASTRAL-Pro and ASTRID-multi are more reliably accurate than FastMuLRFS (with a small advantage to ASTRID-multi), and that ASTRID-multi is often faster than ASTRAL-Pro.

Original languageEnglish (US)
Title of host publicationAlgorithms for Computational Biology - 8th International Conference, AlCoB 2021, Proceedings
EditorsCarlos Martín-Vide, Miguel A. Vega-Rodríguez, Travis Wheeler
Number of pages12
ISBN (Print)9783030744311
StatePublished - 2021
Event8th International Conference on Algorithms for Computational Biology, AlCoB 2021 - Missoula, United States
Duration: Jun 7 2021Jun 11 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12715 LNBI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference8th International Conference on Algorithms for Computational Biology, AlCoB 2021
Country/TerritoryUnited States


  • Gene duplication and loss
  • Species-tree inference

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)


Dive into the research topics of 'Comparing Methods for Species Tree Estimation with Gene Duplication and Loss'. Together they form a unique fingerprint.

Cite this