IQ-TREE

Efficient and versatile phylogenomic software by maximum likelihood http://www.iqtree.org

Introduction

The IQ-TREE software was created as the successor of IQPNNI and TREE-PUZZLE (thus the name IQ-TREE). IQ-TREE was motivated by the rapid accumulation of phylogenomic data, leading to a need for efficient phylogenomic software that can handle a large amount of data and provide more complex models of sequence evolution. To this end, IQ-TREE can utilize multicore computers and distributed parallel computing to speed up the analysis. IQ-TREE automatically performs checkpointing to resume an interrupted analysis.

As input IQ-TREE accepts all common sequence alignment formats including PHYLIP, FASTA, Nexus, Clustal and MSF. As output IQ-TREE will write a self-readable report file (name suffix .iqtree), a NEWICK tree file (.treefile) which can be visualized by tree viewer programs such as FigTree, Dendroscope or iTOL.

Key features of IQ-TREE

• Efficient search algorithm: Fast and effective stochastic algorithm to reconstruct phylogenetic trees by maximum likelihood. IQ-TREE compares favorably to RAxML and PhyML in terms of likelihood while requiring similar amount of computing time (Nguyen et al., 2015).
• Ultrafast bootstrap: An ultrafast bootstrap approximation (UFBoot) to assess branch supports. UFBoot is 10 to 40 times faster than RAxML rapid bootstrap and obtains less biased support values (Minh et al., 2013).
• Ultrafast model selection: An ultrafast and automatic model selection (ModelFinder) which is 10 to 100 times faster than jModelTest and ProtTest. ModelFinder also finds best-fit partitioning scheme like PartitionFinder (Kalyaanamoorthy et al., 2017).
• Phylogenetic testing: Several fast branch tests like SH-aLRT and aBayes test (Anisimova et al., 2011) and tree topology tests like the approximately unbiased (AU) test (Shimodaira, 2002).

The strength of IQ-TREE is the availability of a wide variety of phylogenetic models:

• Common models: All common substitution models for DNA, protein, codon, binary and morphological data with rate heterogeneity among sites and ascertainment bias correction for e.g. SNP data.
• Partition models: Allowing individual models for different genomic loci (e.g. genes or codon positions), mixed data types, mixed rate heterogeneity types, linked or unlinked branch lengths between partitions.
• Mixture Models: fully customizable mixture models and empirical protein mixture models and.
• Polymorphism-aware models (PoMo): http://www.iqtree.org/doc/Polymorphism-Aware-Models

IQ-TREE web service

For a quick start you can also try the IQ-TREE web server, which performs online computation using a dedicated computing cluster. It is very easy to use with as few as just 3 clicks! Try it out at

http://iqtree.cibiv.univie.ac.at

User support

Please refer to the user documentation and frequently asked questions. If you have further questions, feedback, feature requests, and bug reports, please sign up the following Google group (if not done yet) and post a topic to the

https://groups.google.com/d/forum/iqtree

The average response time is one working day.

Citations

When using ModelFinder please cite:

• S. Kalyaanamoorthy, B.Q. Minh, T.K.F. Wong, A. von Haeseler, L.S. Jermiin (2017) ModelFinder: Fast model selection for accurate phylogenetic estimates. Nat. Methods, 14:587-589. https://doi.org/10.1038/nmeth.4285

When performing tree reconstruction please cite:

• L.-T. Nguyen, H.A. Schmidt, A. von Haeseler, and B.Q. Minh (2015) IQ-TREE: A fast and effective stochastic algorithm for estimating maximum likelihood phylogenies. Mol. Biol. Evol., 32, 268-274. https://doi.org/10.1093/molbev/msu300

For the ultrafast bootstrap (UFBoot) please cite:

• D.T. Hoang, O. Chernomor, A. von Haeseler, B.Q. Minh, and L.S. Vinh (2017) UFBoot2: Improving the ultrafast bootstrap approximation. Mol. Biol. Evol., in press. https://doi.org/10.1093/molbev/msx281

When using posterior mean site frequency model (PMSF) please cite:

• H.C. Wang, B.Q. Minh, S. Susko, A.J. Roger (in press) Modeling site heterogeneity with posterior mean site frequency profiles accelerates accurate phylogenomic estimation. Syst. Biol. https://doi.org/10.1093/sysbio/syx068

When using partition models please cite:

• O. Chernomor, A. von Haeseler, B.Q. Minh (2016) Terrace aware data structure for phylogenomic inference from supermatrices. Syst. Biol., 65:997-1008. https://doi.org/10.1093/sysbio/syw037

When using polymorphism-aware models please cite:

• D. Schrempf, B.Q. Minh, N. De Maio, A. von Haeseler, C. Kosiol (2016) Reversible polymorphism-aware phylogenetic models and their application to tree inference. J. Theor. Biol., 407:362-370. https://doi.org/10.1016/j.jtbi.2016.07.042

Credits and Acknowledgements

Some parts of the code were taken from the following packages/libraries: Phylogenetic likelihood library, TREE-PUZZLE, BIONJ, Nexus Class Libary, Eigen library, SPRNG library, Zlib library, gzstream library, vectorclass library, GNU scientific library.

IQ-TREE was partially funded by the Austrian Science Fund - FWF (grant no. I 760-B17 from 2012-2015 and and I 2508-B29 from 2016-2019) and the University of Vienna (Initiativkolleg I059-N).